hypothesis on animal behavior

• Mardigian Library
• Subject Guides

PSYC 372: Animal Behavior

• Develop Your Hypothesis
• Online Library Access
• Find Articles
• What is Peer-Review?
• Read & Analyze Scholarly Articles
• Select Your Articles
• Use Your Articles
• Avoid Plagiarism
• Cite in APA Style
• Write Your Annotations
• Write Your Evolution Essay
• Write Your Zoo Report
• Browse Journals

Behavioral Sciences and Women's & Gender Studies Librarian

Developing Your Hypothesis

For your annoated bibliography, you will use peer-reviewed journal articles to learn more about your species for your zoo observations. You will also use your peer-reviewed journal articles to develop a hypothesis for the behavior you observed during your first zoo observation. 

Do some exploratory research: Start with some exploratory research to learn more about your species before you start your first zoo observation. You will use this exploratory research to develop a hypothesis of the behavior you expect to see at your zoo observation, as well as to identify operational definitions of behaviors to be observed and an ethogram (check sheet on which to collect your data).

Record your observations at the zoo for 3 full hours.  Collate your data and run statistical analysis on it to identify patterns of behavior. 

Identify your variables:   Use the information from your exploratory research and your zoo observation to identify a few behaviors that interest you and the possible relationships between them. Write about your observations, referring back to how they relate to what you found in your expoloratory research.

Choose a current research area:  Develop a hypothesis for species behaviors about which articles are continuing to be published. Avoid defunct or little-known areas of research. 

Write about what interests you:  Professors want students to write about research areas that they care about. If you're interested in the species and behaviors you've chosen, it will be more fun for you to do your observations and write them up, and probably more fun for your professor to read them, too.

Ask Professor Dolins  for feedback on whether the hypothesis you develop is a good hypothesis, one that can be tested.

Picking Your Topic IS Research

Once you've picked a research topic for your paper, it isn't set in stone. It's just an idea that you will test and develop through exploratory research. This exploratory research may guide you into modifying your original idea for a research topic. Watch this video for more info:

• << Previous: Online Library Access
• Next: Find Articles >>
• Last Updated: Sep 12, 2024 7:06 PM
• URL: https://guides.umd.umich.edu/psyc372

Call us at 313-593-5559

Chat with us

Text us: 313-486-5399

Email us your question

• 4901 Evergreen Road Dearborn, MI 48128, USA
• Phone: 313-593-5000
• Maps & Directions
• M+Google Mail
• Emergency Information
• UM-Dearborn Connect
• Wolverine Access

• Log in Username Password Log in Full Login Page / Create Account

The Basics of Animal Behaviour

A short e-Book introducing some of the fundementals of animal behaviour and its research.

6. A Hypothesis

e.g. The age of a penguin will effect its behaviour.

Note that a hypothesis is not a prediction. The researcher is not judged on whether their prediction was correct or not. The research will only be judged on whether they collected the data on the animals behaviour fairly and without bias. This evidence will either support the hypothesis, or reject the hypothesis.

Principles of Animal Behavior, 4th Edition

Fourth Edition

Lee Alan Dugatkin

See resources for instructors . Read the first chapter .

576 pages | 529 color plates, 31 halftones, 3 line drawings | 8 1/2 x 11 | © 2020

Biological Sciences: Anatomy , Behavioral Biology , Evolutionary Biology , Physiology, Biomechanics, and Morphology

• Table of contents
• Author Events

Related Titles

“Up-to-date, highly integrative, and richly illustrated. It thus merits serious consideration by anyone looking for a textbook to support undergraduate offerings in animal behavior or behavioral ecology. . . . Principles of Animal Behavior is comprehensive and readable, summarizing not only what is well documented but more importantly where integrated understanding is lacking, and thus where further research will prove most profitable.”

Animal Behaviour, on the first edition

“The book reveals a richly illustrated panoramic view of animal behavior and, where it can, it also provides examples of the physiological, neurobiological, and molecular genetic mechanisms that may underlie it. . . . Dugatkin’s text . . . can be enjoyed by anyone who has an interest in the beauty of animal behavior. . . . Excellent.”

Times Higher Education, on the second edition

"Dugatkin offers a highly readable overview of the modern field of animal behavior, keeping up with the exponential growth over the last few decades. Instead of the usual choice between inborn and learned behavior, this book offers the reader with a fully integrated view."

Frans de Waal, Emory University | on the third edition

"This is an up-to-date text that shows not only how theory and empirical data are combined, but how they are derived on the ground."

Bernd Heinrich, University of Vermont | on the third edition

Table of Contents

Elephant don.

Caitlin O'Connell

Zebra Stripes

Dr. calhoun’s mousery, the art of being a parasite.

Claude Combes

Be the first to know

Get the latest updates on new releases, special offers, and media highlights when you subscribe to our email lists!

Sign up here for updates about the Press

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 20 October 2020

Deep learning-assisted comparative analysis of animal trajectories with DeepHL

• Takuya Maekawa   ORCID: orcid.org/0000-0002-7227-580X 1 ,
• Kazuya Ohara 1 ,
• Yizhe Zhang 1 ,
• Matasaburo Fukutomi 2 ,
• Sakiko Matsumoto 3 ,
• Kentarou Matsumura 4 ,
• Hisashi Shidara   ORCID: orcid.org/0000-0002-3992-9226 5 ,
• Shuhei J. Yamazaki 6 ,
• Ryusuke Fujisawa 7 ,
• Kaoru Ide 8 ,
• Naohisa Nagaya 9 ,
• Koji Yamazaki 10 ,
• Shinsuke Koike   ORCID: orcid.org/0000-0002-2926-8196 11 ,
• Takahisa Miyatake   ORCID: orcid.org/0000-0002-5476-0676 4 ,
• Koutarou D. Kimura   ORCID: orcid.org/0000-0002-3359-1578 6 , 12 ,
• Hiroto Ogawa   ORCID: orcid.org/0000-0002-4927-9714 5 ,
• Susumu Takahashi 8 &
• Ken Yoda 3  

Nature Communications volume  11 , Article number:  5316 ( 2020 ) Cite this article

12k Accesses

32 Citations

19 Altmetric

Metrics details

• Animal migration
• Data mining
• Machine learning

A comparative analysis of animal behavior (e.g., male vs. female groups) has been widely used to elucidate behavior specific to one group since pre-Darwinian times. However, big data generated by new sensing technologies, e.g., GPS, makes it difficult for them to contrast group differences manually. This study introduces DeepHL, a deep learning-assisted platform for the comparative analysis of animal movement data, i.e., trajectories. This software uses a deep neural network based on an attention mechanism to automatically detect segments in trajectories that are characteristic of one group. It then highlights these segments in visualized trajectories, enabling biologists to focus on these segments, and helps them reveal the underlying meaning of the highlighted segments to facilitate formulating new hypotheses. We tested the platform on a variety of trajectories of worms, insects, mice, bears, and seabirds across a scale from millimeters to hundreds of kilometers, revealing new movement features of these animals.

Similar content being viewed by others

Quantifying behavior to understand the brain

Development of a 3D tracking system for multiple marmosets under free-moving conditions

Identifying behavioral structure from deep variational embeddings of animal motion

Introduction.

Recent advances in sensing technologies such as Global Positioning System (GPS) and computer vision provide “big behavioral data” of animals 1 , 2 , 3 , 4 , 5 . The challenge is how best to capitalize on such data to understand animal behavior, a challenge that has led to many significant cross-disciplinary research projects combining biology and information science 6 , 7 , 8 , 9 . One potentially powerful option involves deep learning artificial intelligence (AI). The recent rapid evolution of this has exceeded the capability of humans in a number of “intelligent” tasks requiring human creativity, including the game of Go 10 , 11 . Because big behavioral data require substantial effort for experts to analyze manually, and because the complexity of the data threatens to blur the capacity for insight, we believe that deep learning-oriented AI is an extremely promising tool to meet complex data challenges. Deep learning and classic machine learning have been used as support tools to quantify animal behavior (e.g., tracking 1 , 12 and behavior recognition 13 ) to reduce the effort involved in manual data labeling by biologists. In contrast, to take deep learning-assisted research one step further, this study leverages deep learning to assist high-level intelligent tasks associated with researchers requiring their insight, for example, the proposal of a hypothesis. Here, we showcase an example of this type of deep learning-assisted research, presenting a computational method that supports the comparative analysis of big behavioral data acquired by, and for, biologists.

Comparative methods have been used by biologists since pre-Darwinian times. Today, with the advent of animal-tracking technologies, comparative analysis, that is, comparison between two groups, for example, experimental vs. control groups and male vs. female groups, is one of the most fundamental approaches to animal behavior analysis. Regarding this, biologists have applied classic knowledge-driven approaches thus far, which are illustrated in the upper portion of Fig.  1 a. In this approach, the biologists typically visually compare huge amounts of time-series movement data, such as hundreds to thousands of trajectories, to identify the behavior that characterizes one group for elucidation, for example, sex-specific movement strategies, which requires substantial effort from the researchers. Then, based on the finding, the biologists design some statistical value computed from the behavioral data that well separates the two groups, which is called as a high-level feature in this study. After that, the biologists validate the finding using the computed high-level features with a statistical test (e.g., testing the significant difference in the high-level features between the male and female groups).

a Difference in research procedures between the conventional and proposed deep learning (DL)-assisted approaches. b Screenshot of the DeepHL web interface comparing the trajectories of a female (left) and male (right) streaked shearwater. Some characteristic segments of the trajectories are highlighted in red; these were detected by our neural network model trained to distinguish between the trajectories of male and female birds. A user can observe that there is something worth investigating in the highlighted segments. In this case, the female trajectory is highlighted when the female bird stays close to the coastline (see  Supplementary Information , Application to the study of seabirds, for more detailed analysis). In contrast, the male trajectory is highlighted when the male bird travels away from the coastline. Note that the small blue and red pins on the maps indicate the starting and terminating points of the trajectories, respectively. The large blue pin on the map moves along the trajectory at a speed proportional to the actual movement speed (Supplementary Movie  1 ). c DeepHL extracts the segments in an input trajectory to which the neural network pays attention when classifying the trajectory by using a time series of the attention values. These segments reveal the importance of each data point. The input trajectory is colored by the time series of the attention values. The range of colors used to color the trajectories is shown on the right of b . d To facilitate a deeper understanding of the implications of the highlighted segments, DeepHL colors trajectories with the values of other sensor data or handcrafted features highly correlated with the attention values; the angle between the vertical axis ( y axis) and a line segment connecting the initial position and each point is used in this example. Points with large angles are focused as shown in the red segments in the male bird trajectory of b . Base map and data copyright OpenStreetMap contributors (License: www.openstreetmap.org/copyright ).

However, this approach possesses the potential risk of researchers overlooking an important high-level feature. This problem is obvious in the big data era. Although trajectory analysis based on classic machine learning has been studied, it still relies on features handcrafted by experienced researchers based on findings discovered by manually browsing a large amount of behavioral data or high-level features designed based on hypotheses formulated in advance, yielding a narrowly focused analysis.

In this study, we present a data-driven approach based on deep learning to support an analysis by biologists, as illustrated in the lower part of Fig.  1 a. Specifically, this study focuses on a comparative analysis, and a deep learning-based method is proposed to help identify the differences between the trajectory data of two groups. With this approach, to extract the high-level features from the trajectory data for a classification of the two groups, we leverage the feature learning capacity of deep learning, that is, learning of the high-level feature extraction processes performed within a deep neural network (DNN), which was originally conducted by experienced researchers. Although a DNN can extract high-level features objectively, unlike a classic approach, a DNN is regarded as a black box, making it difficult to interpret the meaning of the high-level features learned by the DNN, that is, to observe the group differences detected by the network. To address this problem, we developed DeepHL, a free, user-friendly, web-based software, in which an interpretable neural network with multi-scale layer-wise attention 14 is used to elucidate the characteristic segments in the trajectories to which the proposed DNN model focuses on in order to distinguish between the trajectories of the two groups (Fig.  1 b, c). Because this method informs researchers regarding “which parts of the trajectories they should focus on for further analysis,” researchers can save time and effort related to an otherwise manual analysis of huge numbers of trajectories to derive the characteristic segments. In addition, DeepHL finds handcrafted features prepared in advance that are highly correlated with the identified segments to help the researchers consider how best the segments can be explained. Thus, this method facilitates data-driven research for a comparative analysis by supporting knowledge discovery from the data. Figure  1 b shows example outputs of DeepHL when we compare trajectories from male seabirds to those from female seabirds. DeepHL automatically finds trajectory segments characteristic of each sex and then provides visualized trajectories highlighting the relevant segments to researchers. Based on the highlighted trajectories and highly correlated features, biologists develop a new hypothesis related to sex-related difference. Then, the biologists can design a high-level feature to validate the hypothesis.

In this study, we present our analysis on a variety of movement trajectories of worms, insects, and mice in laboratories, and animals in the wild, such as bears and seabirds using DeepHL. Behavioral data of these animals were provided by specialists who have been intensively studying the behavior of these animals by manual analysis and/or classic machine learning. We showed the ability of DeepHL to discover new biological insights that have not been found by the manual analysis or classic machine learning. We believe that DeepHL, a web-based open system, could be the first step to democratize AI for biologists who would otherwise have difficulty setting up computing environments for deep learning.

Here, we briefly introduce the pipeline of the proposed method: (i) DeepHL first trains our proposed network (hereafter called DeepHL-Net) on the trajectory data from two classes. (ii) The attention mechanism in DeepHL-Net then calculates the attention value of each data point in each trajectory for each layer in DeepHL-Net. (iii) Once the attention values are computed, some parts of the trajectories are highlighted by DeepHL using the attention output from a particular layer that is assumed to capture differences in the two classes. To help a user of DeepHL find such a layer (hereinafter, a “discriminator layer”) in DeepHL-Net, DeepHL calculates the score for each layer based on attention outputs from the layer. (iv) DeepHL also supports the user in explaining the meaning of the highlighted segments based on a list of handcrafted features from the trajectories prepared in advance by calculating the correlation between the attention values and each of the handcrafted features (Fig.  1 d).

Before describing our method in detail, herein we provide definitions of the features used in this study. Primitive features are basic features widely used in a locomotion analysis, that is, speed and relative angular speed. Handcrafted features are low-level features handcrafted by researchers, such as acceleration and distance from the initial position, and include the primitive features. High-level features are designed by researchers and characterize a group through a comparative analysis. A high-level feature is computed from the entire trajectory, such as the average movement speed and duration of stay at a feeding location. Although a DNN can also acquire high-level features or concepts, we found it difficult to comprehend these high-level features. A feature calculated in each layer in the DNN is simply called a feature. We explain our method in detail as follows.

Trajectory highlighting in DeepHL

We first explain the method of highlighting trajectories using deep learning. Our method assumes that there are two groups of animals with different properties, that is, class A and class B, and each trajectory belongs either to class A or B. We first convert the time series of the coordinates into time series of movement speed and relative angular speed (Fig.  2 a, b), which are widely used primitive features indicating movement velocity and orientation 15 , 16 , 17 , to achieve position and rotation-invariant trajectory analysis. (For animals that freely move on an agar plate, for example, their absolute coordinates are meaningless.) These features are then fed into DeepHL-Net. DeepHL allows a user to easily input other time series into DeepHL-Net, for example, original coordinates, other primitive features, and other sensor data.

We assume that the trajectories of two classes are given: class A and class B in this example, which corresponds to worms without and with prior odor learning, respectively. a , b Trajectories, that is, a time series of two-dimensional coordinates, are converted into time series of speed and relative angular speed to achieve position- and rotation-invariant analysis. c DeepHL-Net is trained on the time series and then a discriminator layer is found using its attention values. d The discriminator layer outputs a time series of attention values when a trajectory is fed into the trained DeepHL-Net. The length of the time series is identical to that of the time series of speed and relative angular speed. e Each trajectory is colored with its corresponding attention values obtained by the layer. In our system, a large attention value is encoded as red and a small attention value is encoded as yellow, as shown in Fig.  1 b. f Proposed multi-scale layer-wise attention model (DeepHL-Net). The input and output of this model are the time-series primitive features and predicted class, respectively. The model consists of four stacks for the convolutional layers and four stacks for the LSTM layers to extract features at different levels of scale. Blocks labeled “1D Conv and Dropout” and “LSTM and Dropout” indicate the 1D convolutional layer and long short-term memory (LSTM) layer with dropout, respectively. The “Layer-wise attention” block calculates the attention of the outputs of a convolutional/LSTM layer using Eq. ( 1 ). The “MatMul” block multiplies the attention and the outputs of the layer to reflect the segments paid a high level of attention in the classification result. The “Softmax” block indicates the output softmax layer. For more details about the model, see “Methods” section.

DeepHL-Net is designed to classify a trajectory into either class A or B. We train DeepHL-Net on the extracted time series associated with their class labels (Fig.  2 c). Within DeepHL-Net, we identify a discriminator layer that detects characteristic segments, which is detailed later. Because DeepHL-Net is also designed to output the segments in a trajectory to which the discriminator layer pays attention, we color the trajectory using the attention information (Fig.  2 d, e). Figure  2 f shows the architecture of the proposed multi-scale layer-wise attention model (DeepHL-Net) comprising eight stacks of 1D convolutional or long short-term memory (LSTM) layers. Because different filter sizes are used in different convolutional stacks, these stacks are designed to extract features at different levels of scale. In addition, the 1D convolutional layers (orange-colored blocks in Fig.  2 f) extract short-term features. In contrast, the LSTM layers (pink-colored blocks in Fig.  2 f) tend to extract features reflecting long-term dependencies. Furthermore, more abstract features tend to be extracted in deeper layers in each stack. Therefore, the model is designed so that the layers extract features at different levels of temporal scale to classify trajectories. To elucidate which segments of the trajectories are considered to be important by each layer, we introduce an attention mechanism 14 into the model. As shown in Fig.  2 f, the outputs of each 1D convolutional/LSTM layer for an input trajectory are used to compute attention as follows:

Here, \({\bf{a}}\in {{\mathbb{R}}}^{1\times {l}_{\mathrm{MAX}}}\) , which shows the importance (i.e., attention) of each data point in the trajectory and is also used to color the trajectory, is an attention vector that has the same length as the trajectory, where l MAX is the maximum length of the input trajectories. Matrix \(Z\in {{\mathbb{R}}}^{{l}_{\mathrm{MAX}}\times N}\) is an output matrix of the 1D convolutional/LSTM layer, where N is the number of nodes in the convolutional/LSTM layer. Finally, \({W}_{\mathrm{a}}\in {{\mathbb{R}}}^{1\times N}\) and \({b}_{\mathrm{a}}\in {{\mathbb{R}}}^{1\times {l}_{\mathrm{MAX}}}\) are the weight matrix and bias, respectively. The softmax function ensures all the output values sum to 1, and the tanh function limits the output value of its input to a value between  −1 and 1. Equation ( 1 ) is implemented as an artificial neuron in DeepHL-Net (“layer-wise attention”; aqua-colored blocks in Fig.  2 f). The attention is multiplied by the outputs of the 1D convolutional/LSTM layer to contrast the segments to which the layer pays attention (“MatMul”; khaki-colored blocks in Fig.  2 f). The multiplied outputs of all layers are concatenated and then used to output an estimate, that is, class A or B, in a densely connected output layer using the softmax function, that is, the final layer in DeepHL-Net (green-colored block in Fig.  2 f). As mentioned above, our model is designed to calculate attention information at different levels of scale (see “Methods” section for more details about the model).

Comparative analysis using DeepHL

A user of DeepHL discovers knowledge using a web page that displays highlighted trajectories (Fig.  1 b). We explain the usage of DeepHL through an analysis of the roundworm Caenorhabditis elegans , which is commonly used as a model animal in neuroscience to understand how learning modulates behavior 18 , 19 . Previous studies revealed that worms learn prior experience of the repulsive odor 2-nonanone in dopamine-dependent manner 20 , 21 : the worms preexposed to the odor migrate further away from the odor source more efficiently than naive worms do. Interestingly, the average speeds of the worms with or without odor learning are not significantly different, suggesting that the preexposed worms avoid the odor more efficiently. To comprehensively determine the high-level behavioral features characteristic of the repulsive odor learning, we compared the trajectories of naive worms (control class; 163 trajectories) to those of worms preexposed to the odor (preexposed class; 162 trajectories) using DeepHL. The positions of each worm’s centroid on a 9-cm agar plate were recorded at 1 Hz for 600 s (Fig.  3 a; see “Methods” and Supplementary Table  2 ). DeepHL-Net was trained on a multivariate time series of primitive features that DeepHL automatically extracts from the time series of trajectories (see “Methods,” Supplementary Information , Algorithm, and Supplementary Table  1 ). Here, the classification accuracy of the trained DeepHL-Net was 93.9% (see “Methods”), indicating that DeepHL-Net was properly trained. When the accuracy is low, for example, 50%, we can regard such a state as having no differences between the two classes or the training data having certain problems (e.g., an excessively small amount of data).

a The experimental setup (left) for monitoring the worm’s trajectory (right). b Example trajectories of worms colored by attentions of a discriminator layer. Segments of the trajectories corresponding to the run state of the worm are highlighted (in red). c Time series of the moving average of speed (black lines) associated with attention values (colored lines). The upper and lower graphs are obtained from the upper and lower trajectories shown in b , respectively. d Histograms showing the distributions of the moving variance of speed for each time slice within the highlighted trajectory segments. e Frequency analysis of the velocity of a preexposed or control worm. A 128-s-wide sliding window was shifted in 1-sample intervals and the amplitude of each frequency component was obtained from its fast Fourier transform (FFT). The upper and lower spectrograms were, respectively, obtained from the upper and lower trajectories shown in b . f Frequency analysis of the velocity of all the preexposed or control worms computed from entire trajectories. The histograms and box plot show the distributions of the dominant frequency of speed for each time slice. The dominant frequency is the one with the largest amplitude within each window. Significant difference in the dominant frequencies were observed by a generalized linear mixed model (GLMM) with Gaussian distributions ( t  = −6.60; d.f. = 322.8; p  = 1.68 × 10 −10 , effect size( r 2 ) = 0.232; ** p  < 0.01; see “Methods”). The p value is two sided. The box plot shows the 25–75% quartile, with embedded bar representing the median; lower whiskers show Q1 − 1.5 × IQR (Q1: 25% quartile; IQR: interquartile range); upper whiskers show Q3 + 1.5 × IQR (Q3: 75% quartile). Control: n  = 76, 784, preexposed: n  = 75, 750.

In the following, we explain the process of knowledge discovery using the functions of DeepHL:

1. Screening layers: Because DeepHL-Net comprises several layers, DeepHL helps the user find discriminator layers by computing a score for each layer using the following criteria:

A discriminator layer should pay attention only to a portion of a trajectory. Technically speaking, an attention vector from the discriminator layer should have large values within limited segments. When the attention values are identical throughout the entire trajectory, the user cannot determine which part of the trajectory is characteristic of the class of interest.

It is desirable that the way attention is paid to the segments of trajectories belonging to one class by the layer is different from that for another class. Technically speaking, a distribution of attention values using the layer for one class should be different from that for another class. For example, when the layer exhibits large attention values to segments in trajectories belonging to only one class, the user can easily understand that these segments are characteristic of that class.

See “Methods” section for an equation to calculate the score. The DeepHL web interface provides a ranking of the layers based on the calculated scores, enabling the user to easily find high-scoring layers, which can provide an insightful highlight of the trajectory.

2. Showing colored trajectories: In this stage, the user compares trajectories colored by the identified discriminator layer. In the example of Fig.  3 b (colored by a discriminator layer with the highest score), only the relatively straight segments of the trajectories are highlighted in red. In contrast, the layer does not pay attention to segments representing more complex movement (yellow segments). The straight and complex movements reflect the two major behavioral states of the worms: “run” and “pirouette” 15 , 19 . DeepHL found that the run behavior of the preexposed class differs from that of the control class.

Note that, because the number of trajectories to be analyzed is large in many cases, DeepHL has a function for screening the trajectories when the user attempts to show highlighted trajectories by a discriminator layer. Especially when we deal with the trajectories of wild animals, not all trajectories include segments characteristic of a specific class. Therefore, DeepHL computes a score of each trajectory, enabling the user to focus mainly on, for example, trajectories with “female-like” segments. The score is calculated as V ( a ), where a is a time series of attention of the trajectory, to find trajectories with characteristic segments. When the variance value is small, this indicates that the layer does not pay attention to particular segments in the trajectory. In addition, the DeepHL web interface provides a classification result for each trajectory, permitting the user to ignore misclassified trajectories when the user browses trajectories.

3. Understanding meaning of highlights: DeepHL provides two functions to help the user understand the reason why a segment attracts attention using a discriminator layer. The first function provides the correlation between the time series of attention values and each of computed handcrafted features prepared in advance (or other sensor data). This reveals which handcrafted feature is related to the attention of the layer (Supplementary Table  1 ). The second function provides the difference in distributions of each handcrafted feature among the two classes within highlighted segments. This reveals which handcrafted feature has different distributions among the two classes within highlighted segments ( Supplementary Information , Algorithm).

Note that these handcrafted features are intended to help interpret the meaning of the attention of DeepHL-Net and that the handcrafted features do not always completely explain the meaning of the attention. As shown in the animal studies below, the biologists understand the meaning of such attention and then manually design interpretable high-level features, which are used in statistical tests, with the help of the functions.

In the worm example, the absolute correlation coefficient between the attention values of the layer and the moving average of the worm speed is the highest among all handcrafted features (Supplementary Table  3 ). Therefore, we then focus on the speed of the worms. Figure  3 c shows the moving averages of speed associated with the attention values. Here, we can employ the second function to reveal the difference in speed between preexposed and control worms within highlighted segments. Interestingly, the difference in distributions of speed itself between preexposed worms and control worms within highlighted segments is smaller than that of the moving variance of speed (0.22 vs. 0.25; see  Supplementary Information , Algorithm for detailed description about difference computation). DeepHL also provides a graph of the distributions as shown in Fig.  3 d. The graph indicates that the changes in speed of preexposed worms are larger than those of control worms. As shown in the graph related to a control worm (Fig.  3 c, upper panel), we can see that, when attention values are high (colored line), the speed indicates tiny high-frequency changes (black line). In contrast, in one typical example of a preexposed worm (Fig.  3 c, lower panel), when attention values are high, the worm accelerates substantially and maintains a high speed, resulting in large low-frequency changes in speed as well as large moving variance of speed. DeepHL seems to detect the low-frequency changes in the speed of the preexposed worms as a characteristic behavior of the preexposed worms. Consistently, the lower frequency components of the speed of the preexposed worms are more dominant than those of the control worms (Fig.  3 e, f). These results suggest that, in worm odor avoidance behavior, two states for periodic changes in velocity—long-term changes with a peak at 2/128 Hz (i.e., 0.016 Hz; 64 s cycle) and short-term change that peak at 9/128 Hz (i.e., 0.07 Hz; 14.2 s cycle)—exist (Fig.  3 f), and that learning modulates the ratio between these two states to avoid odors efficiently. It is reasonable to speculate that maintaining a high speed (resulting in long-term speed changes) only during the run state contributes to efficient odor–source avoidance behavior. Note that these results were not predicted before this analysis because the average velocities of preexposed and control worms are essentially similar 20 , 22 . The biological significance of the worm and other animal analyses are described in the Supplementary Information.

Application to the study of mice

To test general applicability of DeepHL, we compared the behavioral patterns of normal and Parkinson’s disease (PD) mice freely moving in an open field (Fig.  4 a). Although the primary cause of PD is considered to be the loss of dopaminergic inputs to the striatum, the type of motor symptoms it induces remains unclear. Neurotoxic lesion animal models of PD have been utilized to elucidate the neuronal mechanisms underlying PD. However, in such models, the degree of dopaminergic cell loss can only be established post mortem. To estimate the degree of cell loss before death, several behavioral tests have been developed 23 , 24 . For instance, frequencies of ambulation, immobility, or fine movement epochs in open-field tests are evaluated. We compared normal mice to PD mice using DeepHL to discover a new high-level behavior feature. The classification accuracy for the mouse dataset is 74.7% (see “Methods” for further details). Figure  4 b shows typical examples of trajectories highlighted using a discriminator layer. Segments of the normal mouse trajectory that are far away from the initial position are highlighted in red (see also Fig.  4 d). In addition, DeepHL indicated that the attention values highly correlate with the straight-line distances from the initial position (highest; Supplementary Table  3 ). As shown in Fig.  4 c, when straight-line distances from the initial position exhibit high values, the attention values also increase. This result indicates that the behavior of visiting locations far away from the initial position is characteristic of normal mice.

a Experimental apparatus and lesion protocol. b Example trajectories of mice colored by attention of the discriminator layer. The upper one is a trajectory of a normal mouse and the lower one is a trajectory of a 6-hydroxydopamine (OHDA) lesion mouse model of Parkinson’s disease (PD). The upper trajectory shows that when the mouse is far away from the initial position, the layer pays attention to the corresponding segments (red segments). c A time series of the straight-line distance from the initial position (black lines) associated with attention values (colored lines). d A screenshot of DeepHL for comparing multiple trajectories colored by the discriminator layer at a glance (normal mice). A blue balloon shows the initial position of each trajectory. e – g Average movement speed during ambulation periods, average movement speed during fine movement periods, and average maximum distance within a  ±60-s window in a session of normal and PD mice for each entire trajectory (see “Methods”). Significant differences between normal and PD mice were observed for all three features (Wilcoxon rank-sum test, p  = 3.486 × 10 −5 , p  = 5.869 × 10 −4 , p  = 2.666 × 10 −4 ; ** p  < 0.01; n  = 22 original 10-min trajectories from normal; n  = 30 original 10-min trajectories from PD). The p values are two sided. The edges of the box plot correspond to 95% confidence intervals, the embedded bar represents the median, and whiskers show minimum and maximum values. Dots show values for individual sessions.

To investigate the usefulness of this finding in terms of PD mouse detection, we designed a new high-level movement feature based on it: the maximum straight-line distance within a  ±60 s window. We compare its performance to the performances of existing high-level movement features, that is, ambulation and fine movement speeds (Fig.  4 e–g). We compute these three feature values for each entire trajectory and then evaluated the features using information gain 25 , which is used to evaluate classification features. A larger value of information gain indicates better classification performance. While the ambulation speed, fine movement speed, and maximum distance all exhibited statistical differences between the normal and PD groups, their information gains were 0.269, 0.184, and 0.287, respectively, indicating that the maximum distance is more useful for evaluating PD symptoms than conventional measures.

The results suggest that normal mice prefer exploring unvisited locations. This feature strongly relates to the straight-line distance from the initial position and differs from widely used existing high-level movement features based on speed. It is well known that rodents such as mice and rats spontaneously prefer to explore an environment, particularly in novel places. Thus, DeepHL may have revealed that the abnormal behavior of PD mice hinders such spontaneous behavioral traits.

Application to the study of red flour beetles

In addition to the PD and normal mice, DeepHL was used to detect dopamine-dependent differences in the trajectories of insects. Tonic immobility (TI), sometime called as “thanatosis” or “death-feigning,” is an antipredator behavior of many animals 26 , 27 . Miyatake et al. 28 performed a two-way artificial selection for the duration of TI, and established the strains with short (S strain) and long (L strain) duration of TI in the red flour beetle, Tribolium castaneum . Tribolium castaneum is an insect model species for which all the genomes are already known 29 . The S strain showed significantly higher levels of brain dopamine expression and a higher locomotor activity than those of the L strain 30 . In the present study, we analyzed 419 walking trails collected from S- and L-strain beetles on a treadmill using DeepHL (Fig.  5 a and Supplementary Table  2 ).

a Experimental apparatus (treadmill). b Example trajectories of the red flour beetles colored by the attention values of a discriminator layer. The upper one is a trajectory of the L-strain (long-strain) beetle and the lower one is a trajectory of the S-strain (short-strain) beetle. These trajectories show that segments corresponding to orientation change are highlighted. c Time series of the angle from the initial position (black lines) associated with attention values (colored lines). The upper and lower graphs are obtained from the upper and lower trajectories shown in b , respectively. The lower graph shows that the attention values have large positive values just before the angle increases. d Other trajectories belonging to the S-strain class colored by attention of the discriminator layer. e We assume a circle centered at each point ( p ) on a trajectory with radius r (100 mm) and obtain points n and m where the trajectory first crosses the circle before/after p . We then compute the angle between a line segment connecting p and n and one connecting p and m , showing the curvature around p . f Angles of the L and S strains. The box plot shows the 25–75% quartile, with embedded bar representing the median; lower whiskers show Q1 − 1.5 × IQR (Q1: 25% quartile; IQR: interquartile range); upper whiskers show the maximum values, that is, π , with the violin plots showing the distributions of data points. Significant difference between the L and S strains was observed using the two-sided ANOVA ( F  = 12.57; d.f. = 1; p  = 0.001; effect size ( η 2 ) = 0.09; ** p  < 0.01; see “Methods”). # of data points for S strain is n  = 185, 884; # of data points for L strain is n  = 219, 497.

The classification accuracy for the beetle dataset is 84.5% (see “Methods” for further details). Figure  5 b shows typical examples of trajectories highlighted using a discriminator layer. The trajectories in Fig.  5 b appear to be highlighted when the beetles turn, which is the characteristic difference between the L and S strains detected by DeepHL. Consistently, the difference in distributions of the angle from the initial position between the S and L strains within highlighted segments is large (0.61). We can clearly see that the turn in the S-strain trajectories is sharp, and we found similar patterns in other trajectories. (See Fig.  5 d, generated by a function of DeepHL that allows the comparison of multiple trajectories at a glance.) Figure  5 c shows the angle from the initial position and attention values used for highlighting trajectories in Fig.  5 b, indicating increases in attention values just before increases in the angle for the S strain.

As shown in Fig.  5 e, we computed an angle of a trajectory segment for each point. Figure  5 f shows the distributions of the angles for the S and L strains (the number of data points for the S strain is 185,884 and the number of data points for the L strain is 219,497). We found that the angle for the S strain is significantly smaller than that for the L strain, indicating that the S-strain beetles walk with more angle changing. This finding related to angle change, which has not been discovered by prior studies 30 , may lead to new hypotheses concerning the survival strategy of the beetles. The L-strain beetles are known to perform death-feigning as an antipredator behavior. In contrast, the S-strain beetles are assumed to select a survival strategy of changing movement directions to escape from predators.

Application to the studies of crickets and animals in the wild

We also employed DeepHL to analyze context-dependent modulation of escape behavior in field crickets, Gryllus bimaculatus . Fukutomi et al. 31 , 32 revealed that an acoustic stimulus at high frequency (>10 kHz) preceding an air puff alters crickets’ moving direction in wind-elicited escape behavior, suggesting that the crickets recognize the high-frequency sound as the echolocation signal of bats and change their behaviors in the presence of predators. Here, we adopted DeepHL to compare two groups of escape movement: prestimulated and control (no sound). In this analysis, in addition to the speed and relative angular speed, we input additional sensor data measured using a treadmill, that is, a rotational speed of the body-axis computed from a body-axis angle measured using the treadmill, into DeepHL-Net. Figure  6 a shows typical trajectories colored by the attention values of a discriminator layer. DeepHL shows that the rotational speed of the body axis transiently elevated and peaked earlier in the prestimulated group ( Supplementary Information , Application to the study of crickets; Fig.  6 b), indicating that the sound preceding the air puff provoked the prompt rotational changes of the body axis.

a Trajectories of the prestimulated and control crickets highlighted by DeepHL (see  Supplementary Information , Application to the study of crickets, for more detail). b Time series of rotational speed (black lines) and attention value (colored lines) of the trajectories in a . A discriminator layer pays attention to the difference in the peaks of the rotational speed for the prestimulated and control classes. The high rotational speed was sustained in the control trajectory, which means that the crickets exhibited longer and larger turning movements in the control group. In contrast, the rotational speed transiently elevated and peaked earlier in the prestimulated group. c Trajectories of female and male seabirds highlighted by DeepHL. A discriminator layer pays attention to female trajectories when the locations of the female birds are close to the coastline (see Supplementary Information, Application to the study of seabirds). d Trajectories of female and male bears highlighted by DeepHL. The red segment is characteristic of the male bears, as detected by DeepHL. A discriminator layer pays attention to male trajectories when a male bear has traveled a long distance after/before staying in one place (see Supplementary Information, Application to the study of bears). Base map and data copyright OpenStreetMap contributors (License: www.openstreetmap.org/copyright ).

In addition, we applied DeepHL to the trajectories of wild animals. Figure  6 c shows GPS trajectories of female and male seabirds highlighted by a discriminator layer that pays attention to the migration direction of the birds from their colony. Our investigation revealed that the GPS measurements of the female seabirds are significantly closer to the coastline than those of the male seabirds. In this analysis, in addition to the speed and relative angular speed, we input the absolute coordinates (longitude and latitude) into DeepHL-Net because the absolute coordinates of the specific places such as colonies and feeding sites can affect the behavior of the seabirds. The longitude values were highly correlated with the attention values for the female seabirds. Because the coastline runs north–south, the distance between the coastline and a position is related to the longitude of the position. Therefore, this fact indicates that the behavior of the female seabirds is strongly affected by the distance from the coastline (see Supplementary Information, Application to the study of seabirds). As described above, because we can input an additional time series in addition to the speed and relative angular speed into DeepHL, we can see the effect of the time series on the animal behaviors. Figure  6 d shows the trajectories of female and male bears highlighted by a discriminator layer that pays attention to male trajectories when a male bear travels a long distance after/before it remains in one place. Our investigation revealed that the male bears combined long distance movements with short rests at many locations and the female bears remained in limited locations for a long time (see Supplementary Information, Application to the study of bears).

In the above analysis of the worms, mice, insects, seabirds, and bears, we could discover findings that were not revealed through a manual analysis or classic machine learning. Here, we can easily observe the differences between two groups from the highlighted trajectories of the seabirds and bears. Specifically, it is impossible to observe the relationships between the preferred locations of female seabirds and coastlines without visualization. For the mouse and beetle studies, we also observe the differences between the two groups from the highlighted trajectories. In contrast, it is difficult to find any differences between the two groups related to the worms and crickets by just browsing the highlighted trajectories. Therefore, leveraging both visualization functions and functions to help interpret the meanings of highlights is important to discover knowledge, which is facilitated by the DeepHL web interface (see also Supplementary Information, User guide to DeepHL).

In this study, we demonstrated that DeepHL is able to extract group differences in trajectories for a variety of taxa that operate across scales (for a quantitative evaluation of DeepHL using synthetic trajectory data, see  Supplementary Information , Evaluation with synthetic data). This versatile trajectory analysis was possible because of the useful functions of DeepHL. Furthermore, we confirmed that DeepHL does not require a large number of trajectories to train DeepHL-Net (Supplementary Table  5 ).

Discovering high-level features hidden in temporal dynamics, for example, the frequencies of worm movement speeds and the sustained rotation speed of crickets found with the help of DeepHL, is difficult in classic machine learning without an algorithm specifically designed for each task using prior knowledge gained by manual analysis, which requires substantial effort. In fact, this finding related to the worms has not been discovered by prior study 17 mainly performed by specialists who have been intensively studying the behavior of worms based on a classic approach, which comprehensively extracts 333 handcrafted locomotion features, on the worm data that are also used in our study even though the finding of our study was obtained from the discriminator layer with the highest score. In addition, DeepHL was able to help in finding a prominent mouse movement feature related to exploration, which has not been a focus of prior studies and also obtained from a discriminator layer with the highest score. The discovered movement feature outperformed high-level features found in prior studies in terms of feature importance. This result is surprising because movement features of PD mice have been intensively studied by neuroscientists 23 , 24 . While the movement features of some animals such as seabirds and mice found with the help of DeepHL seem to be simple, the fact that these simple features have not been discovered after many years of research indicates the value of the findings given the difficulties of big behavioral data analysis based on classic approaches. Refer to Supplementary Information , Comparison with classic approaches, for analysis of the six animal species using classic approaches. Whereas the classification accuracy of a classic approach is not extremely different from that of DeepHL, as shown in  Supplementary Information , Comparison with classic approaches, it was difficult to find fine-grained characteristics of animal behaviors by using the classic approach because it employs only high-level features prepared in advance extracted from a whole trajectory.

In this study, we find a discriminator layer that focuses on a part of trajectory. However, it is possible for many attention layers to focus on the full trajectory. In such a case, we can assume that the global features are important for classifying the trajectories. We believe that such global features can be easily identified through classic statistical techniques or manual analysis.

Trajectory data observed from wild animals can include different noises. For example, the trajectory data from bears are noisy because of the forest canopy, as shown in Fig.  6 d. When noises are included in the GPS measurements of both the male and female bears uniformly, we believe that DeepHL-Net can extract useful high-level features from the data. However, such noises can also degrade the classification performance. One possible solution to addressing this problem is to introduce a denoizing autoencoder 33 (e.g., reducing noises during the preprocessing). In addition, the seabird GPS data include few sudden large errors. We can remove such errors by thresholding calculated speed (see  Supplementary Information , Application to the study of seabirds). Moreover, GPS signals can be lost for a moment. However, primitive features used in this study, that is, speed and relative angular speed, are robust against such missing measurements.

There are several deep learning studies related to our work. Endo et al. 34 visualize/generate typical trajectories for taxis using an autoencoder. In addition, several visualization tools for interpreting the behavior of LSTMs have been developed, although these mainly focus on natural language processing 35 , 36 . LSTMs have also been used to predict worm trajectories 37 , although these studies do not focus on comparative analysis. Recent deep learning studies have also employed attention mechanisms to visualize distinguishing features 38 , 39 . Attention mechanisms have also been actively studied in the computer vision field. Xu et al. 40 generated captions for an image by leveraging the attention of the input image to identify an important region in the image and generate each word. Zhang et al. 41 leveraged attention mechanisms to focus on foreground regions to alleviate distractions from the background for image-based salient object detection tasks. Park et al. 42 employed an attention mechanism to identify important regions in an image as well as generate textual descriptions using an LSTM for an image classification task. In the biology domain, Heras et al. 43 leveraged a deep attention network that predicts future turns of a zebrafish in a collective to identify surrounding zebrafish that affect the future turning of the focal zebrafish. Unlike in the above studies, in the present study, deep attention networks have been used to find distinguishing group-specific patterns in the trajectories.

Because the manual analysis of behavioral data is impractical for big behavioral data, we suggest that we are now on the cusp of changing the methods used for big data in biology research. We envision that DeepHL will transform the hitherto standard approach to comparative analysis from a hypothesis-driven approach, which relies on individual experience or manual analysis by researchers, to a data-driven approach. Owing to the useful functions proposed in this study, DeepHL enables researchers to easily extract insightful information hidden within a DNN that is trained on big data. Furthermore, because DeepHL is simply designed to find distinguishing trajectory segments between two groups, it can also be applied to a variety of comparative analyses, for example, old vs. young animals, free-ranging vs. captive animals, animals from different habitats, animals with different life-history stages, food storing vs. non-storing individuals/species, social vs. solitary individuals/species, and specialists vs. generalists. Although many functions of DeepHL are tailored to a trajectory analysis, DeepHL-Net can process any type of time-series data. As a part of a future study, we plan to apply our network to other time series such as sounds emitted by animals.

We believe that these animal behavior analyses contribute not only to biology research, but also to the sustainable development of our society and coexistence with wild animals by understanding animal behavior. In addition, livestock farming has many potential applications of our animal behavior analysis. For example, our method can be applied to identifying characteristic behaviors of disease animals, productive cows, and submissive cows in a social hierarchy. Moreover, because trajectory data are observed from any moving objects, DeepHL is capable of wide application. Specifically, we believe that DeepHL can also be applied to trajectory analyses for humans and automobiles, which can contribute to our society in various aspects: improvement of work efficiency (e.g., analyzing trajectories of workers in logistics centers), healthcare (e.g., comparing between patients and healthy subjects), and eco-safe driving.

DeepHL system architecture

The DeepHL system consists of three server computers. The first one is a web server that receives a trajectory data file from a user and provides analysis results to the user (Intel Xeon E5-2620 v4, 16 cores, 32 GB RAM, Ubuntu 14.04). The second one is a storage server that stores data files and analysis results. The third one is a GPU server that analyzes data provided by the user (Intel Xeon E5-2620 v4, 32 cores, 512 GB RAM, four NVIDIA Quadro P6000, Ubuntu 14.04).  Supplementary Information , Algorithm, provides a complete description of the DeepHL method. DeepHL is accessible on the Internet through http://www-mmde.ist.osaka-u.ac.jp/maekawa/deephl/ .  Supplementary Information , User guide to DeepHL, provides a user guide to DeepHL. In addition, Supplementary Information , Usage of Python-based Software, and Supplementary Software 1 present the Python code of DeepHL.

Preprocessing

An input trajectory is a series of timestamps and X / Y coordinates associated with a class label. To perform position- and rotation-independent analysis, we convert the series into time series of speed and relative angular speed and then standardize them ( Supplementary Information , Algorithm). Note that the absolute coordinates of wild animals, which can relate to the distance from a nest or feeding location, for example, are important in understanding behavior of the animals. Hence, DeepHL allows the original coordinates to be input to DeepHL-Net along with the speed and relative angular speed. In addition, other biological time-series sensor data measured by the user can be fed into DeepHL-Net when these time-series data are included in a data file uploaded by the user. For example, a time series of the heading direction of animals obtained from digital compasses can be useful for behavior understanding. Moreover, primitive features usually used in trajectory analysis can be easily fed into DeepHL-Net. DeepHL automatically computes the travel distance from the initial position, the straight-line distance from the initial position, and the angle from the initial position (Supplementary Table  1 ) as primitive features. Using the web interface of DeepHL, the user can easily select primitive features and other sensor data to be fed into DeepHL-Net ( Supplementary Information , User guide to DeepHL). See  Supplementary Information , Effect of input features, for effects of input features on classification accuracy. Normally, the inputs of DeepHL-Net are two-dimensional time series, that is, speed and relative angular speed. When we input an additional time series (such as the original coordinates) into DeepHL-Net, the additional time series are added as additional dimensions of the inputs.

Multi-scale layer-wise attention model (DeepHL-Net)

Here, we explain DeepHL-Net shown in Fig.  2 f in detail. The input of the model is a time series of primitive features, that is, an l MAX  ×  N f matrix, where l MAX is the maximum length of the input trajectories and N f is the dimensionality of the time series, that is, the number of the primitive features. Because the lengths of observed trajectories are not identical to each other in many cases, we fill in missing elements in the matrix with  −1.0 and mask them when we train DeepHL-Net. In each 1D convolutional layer of the convolutional stacks, we extract features by convolving input features through the time dimension using a filter with a width (kernel size) of F t . We use different filter widths in the four convolutional stacks (3%, 6%, 9%, and 12% of l MAX ) to extract features at different levels of scale. We use a stride (step size) of one sample in terms of the time axis. We also use padding to allow the outputs of a layer to have the same length as the layer inputs. In addition, to reduce an overfitting, we employ a dropout, which is a simple regularization technique in which randomly selected neurons are dropped during training 44 . The dropout rate used in this study is 0.5.

In each LSTM layer of the LSTM stacks, we extract features considering the long-term dependencies of the input features. LSTM is a recurrent neural network architecture with memory cells, and it permits us to learn temporal relationships over a long time scale. LSTM learns long-term dependencies by employing memory cells that hold past information, updating the cell state using write, read, and reset operations with input, output, and forget gates (see  Supplementary Information , Algorithm). In addition, we employ dropout to reduce overfitting. The attention information of each layer is computed by using Eq. ( 1 ), and then it is multiplied by the layer output. Here, the softmax and tanh functions in Eq. ( 1 ) are defined as follows:

Note that parameters in Eq. ( 1 ) for each layer, that is, W a and b a , as well as parameters in the convolutional and LSTM layers are estimated during the network training phase. Here, we introduced the tanh activation function into Eq. ( 1 ) to smooth out the output attention values. When an outlying large value is included in W a Z T  +  b a at time t , attention values other than time t become extremely small without using the tanh function. When we visualize a trajectory using such attention values, only a single data point is colored in red, making it difficult for a user to identify important segments.

Training and testing of DeepHL-Net

The DeepHL user can select the parameters of DeepHL-Net used in the analysis, that is, the number of convolutional/LSTM layers and the number of neurons in each layer (default: four layers with 16 neurons). Then, DeepHL-Net is trained on 80% of randomly selected trajectories to minimize the binary classification error of the training data, employing backpropagation based on Adam 45 ( Supplementary Information , Algorithm). (Note that each trajectory has a class label for binary classification.) Then, the trained DeepHL-Net is tested using the remaining 20% of trajectories to compute the classification accuracy, providing an indication of the degree of difference between the two classes.

Computing the score of each layer

To screen the layers in DeepHL-Net, we compute a score for each layer according to Eq. ( 4 )

Here, \({A}_{i,{C}_{\mathrm{A}}}\) is a set of attention vectors calculated from trajectories belonging to class A using the i th layer. In addition, \({A}_{i,{C}_{\mathrm{B}}}\) is a set of attention vectors calculated from trajectories belonging to class B using the i th layer. As mentioned in the main text, an attention vector from a discriminator layer should have large values within limited segments. Therefore, \({s}_{\mathrm{fc}}({A}_{i,{C}_{\mathrm{A}}},{A}_{i,{C}_{\mathrm{B}}})\) in Eq. ( 4 ) calculates the averaged variance of the attention values normalized by the average length of the trajectories, as described in Eq. ( 5 ). When the layer focuses on a part of a trajectory, the variance increases

Note that V ( ⋅ ) calculates the variance and l ( ⋅ ) calculates the average length of the trajectories. We take the square root of the average variance to derive the average standard deviation. Using \(l({A}_{i,{C}_{\mathrm{A}}}\cup {A}_{i,{C}_{\mathrm{B}}})\) , which calculates the average length of \({A}_{i,{C}_{\mathrm{A}}}\cup {A}_{i,{C}_{\mathrm{B}}}\) , we normalize the computed variance. Because the softmax function in Eq. ( 1 ) ensures that all values sum to 1, resulting in a larger variance for longer trajectories, we normalize the average variance using the average length.

In addition, as mentioned in the main text, the distribution of attention values by the layer for one class should be different from that for another class. Therefore, \({s}_{\mathrm{it}}({A}_{i,{C}_{\mathrm{A}}},{A}_{i,{C}_{\mathrm{B}}})\) calculates the difference between the distributions of the attention values of classes A and B as follows:

Here, h ( ⋅ ) calculates a normalized histogram of attention with 200 bins, and Intersect( ⋅ , ⋅ ) calculates the area overlap between two histograms, and is described as follows:

where H 1 ( i ) shows the normalized frequency of the i th bin of histogram H 1 . As described in Eq. ( 4 ), the final score is calculated as the sum of the two scores of \({s}_{\mathrm{fc}}({A}_{i,{C}_{\mathrm{A}}},{A}_{i,{C}_{\mathrm{B}}})\) and \({s}_{\mathrm{it}}({A}_{i,{C}_{\mathrm{A}}},{A}_{i,{C}_{\mathrm{B}}})\) .

Here, \({s}_{\mathrm{fc}}({A}_{i,{C}_{\mathrm{A}}},{A}_{i,{C}_{\mathrm{B}}})\) in Eq. ( 4 ) is used to find a layer that focuses only on a portion of a trajectory. Owing to the term, only a small important portion of trajectories is highlighted in many cases, as shown in Figs.  3 , 5 , and 6 , especially for the trajectories of beetles. However, substantial portions of several trajectories of the normal mice are highlighted, as shown in Fig.  4 d. Because the characteristics of the normal mouse trajectories are the distance from the initial position, the segments in the trajectories far from the initial position are highlighted.

Computing the correlation between attention values and handcrafted features

To help the user understand the meaning of the highlights, DeepHL automatically computes the Pearson correlation coefficients between the attention values of each layer and handcrafted features computed by DeepHL, as shown in Supplementary Table  1 . In addition, the correlation coefficients with sensor data and handcrafted features included in a trajectory data file are automatically computed. Computing the correlation with environmental sensor data can reveal the relationship between a behavior and environmental conditions. If a specific behavior is exhibited only when the temperature is high, for example, we can infer that the behavior relates to the high temperature condition. Furthermore, DeepHL automatically computes the moving average, moving variance, and derivative of each of the above features/sensor data, and then computes the correlation coefficients with the attention values, which are presented to the user (Supplementary Fig.  1 ).

Computing the difference between distributions of each handcrafted feature for the two classes within highlighted segments

To help the user understand the meaning of the highlights, DeepHL automatically computes the difference between distributions of each handcrafted feature for two classes within highlighted segments. The difference is computed as follows:

Here, \({F}_{j,{C}_{\mathrm{A}}}\) is a set of time series of the j th handcrafted feature calculated from trajectories belonging to class A. In addition, m ( ⋅ , ⋅ ) is a masking function that extracts feature values within highlighted segments. Because the softmax function in each attention layer ensures that all attention values in a sum of 1, we consider an attention value larger than c /(# time slices) as a potential attended value ( c  = 1.2 in our implementation).

Data acquisition of worms

Data acquisition was performed according to Yamazoe-Umemoto et al. 22 . In brief, several worms were placed in the center of an agar plate in a 9-cm Petri dish, 30% 2-nonanone (v/v, EtOH) was spotted on the left side of the plate, which was covered by a lid and placed on the bench upside down. Then, the images of the plate were captured with a high-resolution USB camera for 12 min at 1 Hz. Because the worms do not exhibit odor avoidance behavior during the first 2 min because of the rapid increase in odor concentration 46 , the data for the following 10 min (i.e., 600 s) was used. From the images, individual worms were identified and the position of the centroid was recorded by an image processing software Move-tr/2D (v. 8.31; Library Inc., Japan). The number of recorded trajectories is 325 (Supplementary Table  2 ). The comparison was between the naive worms (control class) and the worms after preexposure to the odor (preexposed class).

DeepHL analysis of worms

A multivariate time series of movement speed, relative angular speed, distances from the initial position, and angle from the initial position extracted from the time series of trajectories was fed into DeepHL-Net, yielding a binary classification accuracy of 93.9%, where 20% of the data are used as test data. The discriminator layer used in this investigation has the highest score of all layers. As shown in Fig.  3 d, which was calculated from the moving variance of the speed within highlighted segments, we can state that the changes in the speed of preexposed worms is larger than those of control worms. Figure  3 e shows spectrograms of the speed calculated from entire trajectories (Fig.  3 c) with a 128-s wide sliding window shifted in 1-sample intervals. In addition, Fig.  3 f shows histograms of the dominant frequency of speed calculated from entire trajectories using the 128-s wide sliding window shifted in 1-sample intervals. These results also indicate the difference in the frequency of speed between the preexposed and control worms. Our investigation revealed that the dominant frequency of speed significantly differs between the preexposed and control worms using GLMM with Gaussian distributions ( t  = −6.60; d.f. = 322.8; p  = 1.68 × 10 −10 , effect size( r 2 ) = 0.232). The p value is two sided. Individual factors were treated as random effects. The number of data points for the control class is n  = 76, 784 and that for the preexposed class is n  = 75, 750. We used GLMM with Gaussian distributions because the objective variable has a continuous value and we used the lmerTest package (v. 2.0–36) of R (v. 3.4.3) for the analysis.

Data acquisition of mice

We collected 52 trajectories of normal mice and unilateral 6-hydroxydopamine (OHDA) lesion mouse models of PD while they freely moved for 10 min in an open field (60 × 55 cm 2 , wall height = 20 cm; normal: 22, PD: 30). The trajectories were detected by the animal’s head position, which was captured by an overhead digital video camera (60 fps). Two sets of small red and green light-emitting diodes were mounted above the animal’s head so that it could be located in each frame. Custom softwares based on Matlab (R2018b, Mathworks, MA, USA) and LabVIEW (Labview 2018, National Instruments, TX, USA) were used for tracking. We then created 30-s segments by splitting each trajectory because training a DNN requires a number of trajectories. We used 966 segments in total (normal: 374, PD: 592) collected from nine C57BL/6J mice (normal: 5, PD: 4). Note that we excluded 30-s segments that contain no movements of a mouse.

DeepHL analysis of mice

Movement speed, relative angular speed, travel distances, straight-line and travel distances from the initial position, and angle from the initial position were fed into our model. The accuracy for the binary classification of normal and 6-OHDA model mice was 74.7%, where 20% of the data are used as test data. The score of the discriminator layer was the highest of all LSTM layers and the sixth highest of all layers. Our investigation revealed that the behavior of visiting locations far away from the initial position can be characteristic of normal mice.

To evaluate PD symptoms from animal behaviors, previous studies have exclusively focused on the movement speed of animals in the open-field tests (frequency and bout duration of ambulation as well as immobility or fine movement) because typical symptoms in the animal model of PD are thought to be slowness of movement and a paucity of spontaneous movements. As shown in Fig.  4 e–g, we found significant differences in average movement speed during ambulation periods, average movement speed during fine movement periods, and average maximum distance within a ±60-s window in a session. These differences were derived from the findings of DeepHL using the two-sided Wilcoxon rank-sum test ( W  = 544, p  = 3.486 × 10 −5 , effect size (Cliff’s delta) = −0.648; W  = 511, p  = 5.869 × 10 −4 , effect size (Cliff’s delta) = −0.548; W  = 521, p  = 2.666 × 10 −4 , effect size (Cliff’s delta) = −0.579). The 95% confidence intervals are [1.222, 3.481], [0.139, 0.468], and [13.726, 43.175], respectively. We used the exactRankTests package (v. 0.8–29) of R (v. 3.2.3). Note that these behavioral features are extracted from original 10-min trajectories.

The maximum distance, which was derived from a finding of DeepHL, is more useful for evaluating the PD symptoms than conventional measures based on the movement speed. Note that the new feature is designed based on an insight drawn from an analysis by deep learning. These results suggest that DeepHL helps find a novel measure not directly linked to the movement speed, that is, a straight-line distance within a certain time window. When the aim of an animal is to visit all locations in an area, the travel distance over a short duration commonly becomes longer. Besides, it is well known that rodents, including mice and rats, spontaneously prefer to explore an environment, particularly in novel places. Thus, DeepHL may capture the fact that the abnormal behavior of the 6-OHDA lesion model of PD hinders such spontaneous behavioral traits of normal mice. Indeed, the 6-OHDA lesion mouse model appears to remain in the same place. Although this hypothesis should be verified based on the causality between behavioral traits and neural activity patterns underlying PD symptoms using neuronal recording together with its optogenetic manipulation in the basal ganglia and motor cortex 23 , it is beyond the scope of this study.

Behavioral features of mice

According to Kravitz et al. 23 , ambulation was defined as periods when the velocity of the animal’s center point averaged >2 cm/s for at least 0.5 s. Immobility was defined as continuous periods of time during which the average change of the trajectory was <1 cm for at least 1 s. Fine movement was defined as any movement that was not ambulation or immobility. Maximum travel distance within a  ±60-s window was defined as the maximum straight-line distance between the center of the window and each point within the window. Note that each feature value is computed for each entire 10-min trajectory.

6-OHDA injection of mice

Under isoflurane anesthesia, 6-OHDA (4 mg/ml; Sigma) was injected through the implanted cannulae (AP −1.2 mm, ML 1.1 mm, DV 5.0 mm, 2 μl). Animals were allowed to recover for at least 1 week before post-lesion behavioral testing.

Histological verification of dopaminergic cell loss

After the mice were sacrificed by pentobarbital sodium overdose and perfused with formalin, their brains were frozen and cut coronally at 30 μl with a sliding microtome. For immunostaining, sections were divided into six interleaved sets. Immunohistochemistry was performed on the free-floating sections. Sections were pretreated with 3% hydrogen peroxide and incubated overnight with primary antibody mouse anti-tyrosine hydroxylase (1:1000; Millipore). As a secondary antibody, we used biotinylated donkey anti-mouse IgG (1:100; Jackson ImmunoResearch Inc.), followed by incubation with avidin–biotin–peroxidase complex solution (1:100; VECTASTAIN Elite ABC STANDARD KIT, Vector Laboratories). The immunoreactivities were visualized by 3-3′ diaminobenzidine tetrahydrochloride (Dojindo Laboratories). The degree of dopaminergic cell loss was estimated by dividing the number of cells manually counted across three sections of the SNc (most rostral, most caudal, and the intermediate between them) of the lesioned hemisphere from that of the non-lesioned hemisphere.

Data acquisition of beetles

In the present study, we analyzed 419 walking trails collected from S- and L-strain beetles freely moving on a treadmill 47 (tracking software: custom software based on OpenCV, v. 2.4.9) using DeepHL (Supplementary Table  2 ). The number of the S-strain (L-strain) beetles is 20, consisting of 10 males and 10 females. The sampling rate of the treadmill is ~14.3 Hz, and the average duration of the trajectories is 52 s.

DeepHL analysis of beetles

In addition to the movement speed and relative angular speed, the distances and angle from the initial position were fed into DeepHL-Net. The classification accuracy for the binary classification between S and L strains was 84.5%, where 20% of the data are used as test data. The score of the discriminator layer in Fig.  5 b was the third highest of all layers. Because the layer seems to focus on turns, we computed an angle of a trajectory segment for each point according to Fig.  5 e. Figure  5 f shows the average angles for the S and L strains (the number of data points for the S strain is 185,884 and the number of data points for the L strain is 219,497). We found that the angle for the S strain is significantly larger than that for the L strain, indicating that beetles of the S-strain beetles walk with more angle changing. Note that we used two-sided analysis of variance (ANOVA) ( F  = 12.57; d.f. = 1; p  = 0.001; effect size( η 2 ) = 0.09). Because multiple data points were computed from each individual beetle’s trajectory, we treated the individuals as a random factor. The 95% confidence interval is [0.05, 0.18]. We used JMP 12.2.0., SAS. This result could indicate the difference in strategies for survival between the S- and L-strain beetles. The L-strain beetles can survive because of their long duration of TI against predators. In contrast, the S-strain beetles attempt to escape from a predator by frequently changing their moving directions.

Previous studies have shown a significantly lower expression level of brain dopamine in the beetles derived from the L strain than those from the S strain 30 . Nishi et al. 48 showed that injection of caffeine, which activates dopamine, decreases the duration of immobility in the L strain of T. castaneum . These phenomena concerning dopamine show an analogy to PD, which alters walking patterns 49 . In many animals, dopamine expression level relates to movement patterns, and a specific trajectory segment pattern of the L strain might be similarly deeply affected. To test this new hypothesis, the relationship between dopamine expression and detailed analysis for walking ability, which should be done apart from the present study, should be examined in the future. In conclusion, the analysis using DeepHL revealed significantly different walking trajectories between beetles from the S and L strains using ANOVA: the S-strain beetles walk with more angle changes along the direction of travel compared to the L-strain beetles.

Ethics statement

The studies on streaked shearwaters, mice, and bears were approved by the Animal Experimental Committees of Nagoya University (streaked shearwaters), the Doshisha University Institutional Animal Care and Use Committees (mice), and the Institutional Animal Care and Use Committee of Tokyo University of Agriculture and Technology (bears), respectively. The research on streaked shearwaters was conducted with permits from the Ministry of the Environment, Japan. All experimental procedures used in the bear research followed the Guidelines Concerning Animal Experimentation of the Tokyo University of Agriculture and Technology and the Mammal Society of Japan. They specify no requirements for the treatment of insects in experiments. Details of animals used in this study are described in  Supplementary Information , Animals.

Reporting summary

Further information on research design is available in the  Nature Research Reporting Summary linked to this article.

Data availability

The dataset of the worms analyzed during the current study is available in the Dryad repository, https://doi.org/10.5061/dryad.37pvmcvf5 , and included in  Supplementary Data 1 . The datasets of the mice, beetles, crickets, and seabirds analyzed during the current study are included in  Supplementary Data 1 . The dataset of the bears are available from the corresponding author upon reasonable request because the release of the bear data can increase the likelihood of poaching and stir up the fear in residents.  Source data are provided with this paper.

Code availability

The source code of DeepHL is distributed as  Supplementary Software 1 . The most recent version of the software is available at https://doi.org/10.5281/zenodo.4023931 50 . The use of the software is exclusively limited to the purpose of undertaking academic, governmental, or not-for-profit research. The DeepHL web system is accessible on the Internet through http://www-mmde.ist.osaka-u.ac.jp/maekawa/deephl/ . We will keep the website operating and freely accessible for the foreseeable future.

Pérez-Escudero, A., Vicente-Page, J., Hinz, R. C., Arganda, S. & de Polavieja, G. G. idTracker: tracking individuals in a group by automatic identification of unmarked animals. Nat. Methods 11 , 743–748 (2014).

Article   PubMed   Google Scholar  

Cagnacci, F., Boitani, L., Powell, R. A. & Boyce, M. S. Animal ecology meets GPS-based radiotelemetry : a perfect storm of opportunities and challenges. Philos. Trans. R. Soc. Ser. B 365 , 2157–2162 (2010).

Article   Google Scholar  

Block, B. A. et al. Tracking apex marine predator movements in a dynamic ocean. Nature 475 , 86–90 (2011).

Article   CAS   PubMed   Google Scholar  

Ropert-Coudert, Y. & Wilson, R. P. Trends and perspectives in animal-attached remote sensing. Front. Ecol. Environ. 3 , 437–444 (2005).

Gomez-Marin, A., Paton, J. J., Kampff, A. R., Costa, R. M. & Mainen, Z. F. Big behavioral data: psychology, ethology and the foundations of neuroscience. Nat. Neurosci. 17 , 1455–1462 (2014).

Zheng, Y. Trajectory data mining: an overview. ACM Trans. Intell. Syst. Technol. 6 , 1–41 (2015).

Wang, R. F. & Spelke, E. S. Human spatial representation: Insights from animals. Trends Cogn. Sci. 6 , 376–382 (2002).

Sugden, A. & Pennisi, E. When to go, where to stop. Science 313 , 775 (2006).

Article   CAS   Google Scholar  

Trullier, O., Wiener, S. I., Berthoz, A. & Meyer, J. Biologically based artificial navigation systems: review and prospects. Prog. Neurobiol. 51 , 483–544 (1997).

Silver, D. et al. Mastering the game of Go with deep neural networks and tree search. Nature 529 , 484–489 (2016).

Article   ADS   CAS   Google Scholar  

Silver, D. et al. Mastering the game of Go without human knowledge. Nature 550 , 354–359 (2017).

Article   ADS   CAS   PubMed   Google Scholar  

Mathis, A. et al. DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nat. Neurosci. 21 , 1281–1289 (2018).

Kabra, M., Robie, A. A., Rivera-Alba, M., Branson, S. & Branson, K. JAABA: interactive machine learning for automatic annotation of animal behavior. Nat. Methods 10 , 64–67 (2013).

Bahdanau, D., Cho, K. & Bengio, Y. Neural machine translation by jointly learning to align and translate. Preprint at https://arxiv.org/abs/1409.0473 (2014).

Pierce-Shimomura, J. T., Morse, T. M. & Lockery, S. R. The fundamental role of pirouettes in Caenorhabditis elegans chemotaxis. J. Neurosci. 19 , 9557–9569 (1999).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Bennison, A. et al. Search and foraging behaviors from movement data: a comparison of methods. Ecol. Evol. 8 , 13–24 (2018).

Yamazaki, S. J. et al. STEFTR: A hybrid versatile method for state estimation and feature extraction from the trajectory of animal behavior. Front. Neurosci. 13 , 626 (2019).

Article   PubMed   PubMed Central   Google Scholar  

Bargmann, C. I. Chemosensation. in WormBook (ed. The C. elegans Research Community) (2006).

Lockery, S. R. The computational worm: spatial orientation and its neuronal basis in C. elegans . Curr. Opin. Neurobiol. 21 , 782–790 (2011).

Kimura, K. D., Fujita, K. & Katsura, I. Enhancement of odor avoidance regulated by dopamine signaling in Caenorhabditis elegans . J. Neurosci. 30 , 16365–16375 (2010).

Kimura, K. D. & Yamazaki, S. J. The data of  ~160 worms during odor avoidance behavior before or after learning. Dryad https://doi.org/10.5061/dryad.37pvmcvf5 (2019).

Yamazoe-Umemoto, A., Fujita, K., Iino, Y., Iwasaki, Y. & Kimura, K. D. Modulation of different behavioral components by neuropeptide and dopamine signalings in non-associative odor learning of Caenorhabditis elegans . Neurosci. Res. 99 , 22–33 (2015).

Kravitz, A. V. et al. Regulation of parkinsonian motor behaviors by optogenetic control of basal ganglia circuitry. Nature 466 , 622–626 (2013).

Article   ADS   Google Scholar  

Boix, J., Padel, T. & Paul, G. A partial lesion model of Parkinson’s disease in mice—characterization of a 6-OHDA-induced medial forebrain bundle lesion. Behav. Brain Res. 284 , 196–206 (2015).

Kullback, S. Information Theory and Statistics , Vol. 1 (Dover Publications, New York, 1959).

Rogers, S. M. & Simpson, S. J. Thanatosis. Curr. Biol. 24 , R1031–R1033 (2014).

Humphreys, R. K. & Ruxton, G. D. A review of thanatosis (death feigning) as an anti-predator behaviour. Behav. Ecol. Sociobiol. 72 , 22 (2018).

Miyatake, T. et al. Is death-feigning adaptive? Heritable variation in fitness difference of death-feigning behaviour. Proc. R. Soc. Ser. B 271 , 2293–2296 (2004).

Tribolium Genome Sequencing Consortium. The genome of the model beetle and pest Tribolium castaneum . Nature 452 , 949–955 (2008).

Miyatake, T. et al. Pleiotropic antipredator strategies, fleeing and feigning death, correlated with dopamine levels in Tribolium castaneum . Anim. Behav. 75 , 113–121 (2008).

Fukutomi, M. & Ogawa, H. Crickets alter wind-elicited escape strategies depending on acoustic context. Sci. Rep. 7 , 15158 (2017).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Fukutomi, M., Someya, M. & Ogawa, H. Auditory modulation of wind-elicited walking behavior in the cricket Gryllus bimaculatus . J. Exp. Biol. 218 , 3968–3977 (2015).

Vincent, P., Larochelle, H., Bengio, Y. & Manzagol, P.-A. Extracting and composing robust features with denoising autoencoders. In the 25th International Conference on Machine Learning , 1096–1103 (2008).

Endo, Y., Toda, H., Nishida, K. & Ikedo, J. Classifying spatial trajectories using representation learning. Int. J. Data Sci. Anal. 2, 107–117 (2016).

Karpathy, A., Johnson, J. & Fei-Fei, L. Visualizing and understanding recurrent networks. Preprint at https://arxiv.org/abs/1506.02078 (2015).

Strobelt, H., Gehrmann, S., Pfister, H. & Rush, A. M. LSTMVis: a tool for visual analysis of hidden state dynamics in recurrent neural networks. IEEE Trans. Vis. Comput. Graph. 24 , 667–676 (2018).

Li, K. et al. Recurrent neural networks with interpretable cells predict and classify worm behaviour. Preprint at bioRxiv 222208 (2017).

Lin, Z. et al. A structured self-attentive sentence embedding. Preprint at https://arxiv.org/abs/1703.03130 (2017).

Vaswani, A. et al. Attention is all you need. In Advances in Neural Information Processing Systems , 5998–6008 (2017).

Xu, K. et al. Show, attend and tell: neural image caption generation with visual attention. In International Conference on Machine Learning , 2048–2057 (2015).

Zhang, X., Wang, T., Qi, J., Lu, H. & Wang, G. Progressive attention guided recurrent network for salient object detection. In IEEE Conference on Computer Vision and Pattern Recognition , 714–722 (2018).

Park, D. H. et al. Multimodal explanations: justifying decisions and pointing to the evidence. In IEEE Conference on Computer Vision and Pattern Recognition , 8779–8788 (2018).

Heras, F. J., Romero-Ferrero, F., Hinz, R. C. & de Polavieja, G. G. Deep attention networks reveal the rules of collective motion in zebrafish. PLoS Comput. Biol. 15 , e1007354 (2019).

Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I. & Salakhutdinov, R. Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15 , 1929–1958 (2014).

MathSciNet   MATH   Google Scholar  

Kingma, D. P. & Ba, J. L. Adam: a method for stochastic optimization. Preprint at https://arxiv.org/abs/1412.6980 . 1–15 (2014).

Tanimoto, Y. et al. Calcium dynamics regulating the timing of decision-making in C. elegans . eLife 6 , e21629 (2017).

Nagaya, N. et al. Anomalous diffusion on the servosphere: a potential tool for detecting inherent organismal movement patterns. PLoS ONE 12 , e0177480 (2017).

Nishi, Y., Sasaki, K. & Miyatake, T. Biogenic amines, caffeine and tonic immobility in Tribolium castaneum . J. Insect Physiol. 56 , 622–628 (2010).

Gelb, D. J., Oliver, E. & Gilman, S. Diagnostic criteria for Parkinson disease. Arch. Neurol. 56 , 33–39 (1999).

Maekawa, T. et al. Deep learning-assisted comparative analysis of animal trajectories with DeepHL. GitHub https://doi.org/10.5281/zenodo.4023931 (2020).

Download references

Acknowledgements

We thank Rory P. Wilson for comments on the manuscript. We are also grateful to Chinatsu Kozakai, Tomoya Abe, Masahiro Ogawa, and Maki Yamamoto for their field assistance. This work was supported by JSPS Kakenhi JP16H06539, JP16H06545, JP16H06544, JP16H06543, JP16H06541, JP17H05976, and JP17H05971.

Author information

Authors and affiliations.

Graduate School of Information Science and Technology, Osaka University, Osaka, Japan

Takuya Maekawa, Kazuya Ohara & Yizhe Zhang

Graduate School of Life Science, Hokkaido University, Hokkaido, Japan

Matasaburo Fukutomi

Graduate School of Environmental Studies, Nagoya University, Nagoya, Japan

Sakiko Matsumoto & Ken Yoda

Graduate School of Environmental and Life Science, Okayama University, Okayama, Japan

Kentarou Matsumura & Takahisa Miyatake

Department of Biological Sciences, Hokkaido University, Hokkaido, Japan

Hisashi Shidara & Hiroto Ogawa

Graduate School of Science, Osaka University, Osaka, Japan

Shuhei J. Yamazaki & Koutarou D. Kimura

Graduate School of Computer Science and Systems Engineering, Kyushu Institute of Technology, Iizuka, Japan

Ryusuke Fujisawa

Graduate School of Brain Science, Doshisha University, Kyotanabe, Japan

Kaoru Ide & Susumu Takahashi

Department of Intelligent Systems, Kyoto Sangyo University, Kyoto, Japan

Naohisa Nagaya

Department of Forest Science, Tokyo University of Agriculture, Tokyo, Japan

Koji Yamazaki

Graduate School of Agriculture, Tokyo University of Agriculture and Technology, Tokyo, Japan

Shinsuke Koike

Graduate School of Science, Nagoya City University, Nagoya, Japan

Koutarou D. Kimura

You can also search for this author in PubMed   Google Scholar

Contributions

T.Maekawa conceived and directed the study, and performed method design, software implementation, data analysis, and manuscript writing. K.O. and Y.Z. performed method design, software implementation, and data analysis. K.D.K., S.J.Y., K.Yoda, S.T., H.O., H.S., M.F., T.Miyatake, K.M., and S.K. performed data collection, data analysis, and manuscript writing. S.M., R.F., N.N., K.Yamazaki, and K.I. performed data collection and developed data collection devices.

Corresponding author

Correspondence to Takuya Maekawa .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks Gonzalo de Polavieja and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, peer review file, description of additional supplementary files, supplementary data 1, supplementary movie 1, supplementary software 1, reporting summary, source data, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Maekawa, T., Ohara, K., Zhang, Y. et al. Deep learning-assisted comparative analysis of animal trajectories with DeepHL. Nat Commun 11 , 5316 (2020). https://doi.org/10.1038/s41467-020-19105-0

Download citation

Received : 20 November 2019

Accepted : 25 September 2020

Published : 20 October 2020

DOI : https://doi.org/10.1038/s41467-020-19105-0

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Ptds centertrack: pedestrian tracking in dense scenes with re-identification and feature enhancement.

• Jiazheng Wen

Machine Vision and Applications (2024)

Ponds as experimental arenas for studying animal movement: current research and future prospects

• Christer Brönmark
• Gustav Hellström
• Kaj Hulthén

Movement Ecology (2023)

Cross-species behavior analysis with attention-based domain-adversarial deep neural networks

• Takuya Maekawa
• Daiki Higashide
• Susumu Takahashi

Nature Communications (2021)

Unraveling hidden interactions in complex systems with deep learning

• Seungwoong Ha
• Hawoong Jeong

Scientific Reports (2021)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• Search Menu
• Sign in through your institution
• Advance articles
• Editor's Choice
• Virtual Issue archive
• The Pitelka Award
• Author Guidelines
• Submission Site
• Open Access
• Why publish with us?
• About Behavioral Ecology
• About the International Society for Behavioral Ecology
• Editorial Board
• Advertising and Corporate Services
• Journals Career Network
• Self-Archiving Policy
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

Introduction, animal personalities—what are they and should we care, does “personality” provide any conceptual advances, why do we see so much variation in behavior, life-history evolution, trade-offs, and pace of life, behavioral genetics and behavioral neuroendocrinology: understanding behavioral variation, final words, does the field of animal personality provide any new insights for behavioral ecology.

Address correspondence to M. Beekman. E-mail: [email protected] .

• Article contents
• Figures & tables
• Supplementary Data

Madeleine Beekman, L. Alex Jordan, Does the field of animal personality provide any new insights for behavioral ecology?, Behavioral Ecology , Volume 28, Issue 3, 01 May-June 2017, Pages 617–623, https://doi.org/10.1093/beheco/arx022

• Permissions Icon Permissions

The field of animal personalities claims to fill a gap in our understanding of animal behavior, because it explicitly studies the adaptive significance of behavioral differences. This is a controversial claim given that the field of behavioral ecology firmly places the study of animal behavior in an evolutionary context. In fact, it is the evolutionary context that differentiates behavioral ecology from ethology and animal behavior, 2 fields that were already concerned with the study of behavior in nonhuman animals. So, if behavioral ecology already takes an evolutionary approach to variation in behavior, we ask what is personality research about exactly? This question is particularly pertinent now the focus of personality research shifts and the field moves away from being mainly descriptive to include quantitative frameworks. As a result, the field has come to borrow heavily from already established fields. In our view, this has resulted in “animal personality” studies becoming nothing more than a rebranding of existing fields of research—fields that are far more solidly grounded and hypothesis driven than the often vague and superficial focus on animal personalities.

“[…] and unless profitable variations do occur, natural selection can do nothing.” Charles Darwin In The Origin Of Species , Chapter Natural Selection

The last decade of animal behavior research has seen the rise of animal personality; a field of study defined as examining individual differences in behavior, or suites of correlated behaviors, that are consistent over time and context. Initially, the field was driven by the perceived need to study correlated behaviors not in isolation, as was allegedly done in behavioral ecology, but together ( Sih et al. 2004a , 2004b ). By incorporating variation among or within individuals into behavioral ecology, the field attempts to elucidate when and where individual plasticity or fixed behaviors might evolve ( Dall et al. 2004 ). The field’s initial goal quickly led to a divergence in the main approaches taken by researchers. On the one hand, a strong theoretical body of literature arose to examine how within- or among-individual differences might evolve ( McElreath and Strimling 2006 ; Wolf et al. 2007 , 2008 , 2011 ; Wolf and McNamara 2012 ). Simultaneously, we saw a rapid increase of mainly descriptive papers pointing to correlations among behaviors and measuring behavioral repeatability (reviewed in Bell et al. 2009 ). Even though in the latter case claims were made that such studies provide novel insights into the existence of behavioral variation, little attempt was made to link observed behaviors to evolutionary theory. Unsurprisingly, this latter approach was met with bemusement among many in the field of behavioral ecology, a field that explicitly places animal behavior in an evolutionary context and, by doing so, deals directly with interindividual differences in the traits of interest. Moreover, there have always been fields directly concerned with understanding both the proximate and ultimate mechanisms underlying variation in behavior—for example, life-history theory, foraging theory, behavioral genetics, game theory, quantitative genetics, reaction norms and behavioral (neuro)endocrinology. The main difference appears to be that the field of personality stresses the importance of behavioral consistency, although the focus on consistency was later relaxed (see further).

An early verbal and often informal criticism of animal personality was that it was simply “putting old wine in new bottles,” and that the field contributed nothing new except a media-friendly hype term ( Crews 2013 ). We do not fully agree with this criticism, as the focus on animal personality has certainly led to advances in the tools available to behavioral ecologists to examine interindividual variation and repeatability of behavioral traits ( Dingemanse, Dochtermann, et al. 2010 ; Mathot et al. 2012 ; Roche et al. 2016 ). In fact, the need to quantify consistent behavioral differences has led to a separate field that has been described as “behavioral ecology of variance components,” which traces its origins to the statistical framework employed in quantitative genetics ( Dingemanse and Dochtermann 2014 ; Araya-Ajoy et al. 2015 ; Westneat et al. 2015 ). Our criticism of the field of animal personality lies in its appropriation of existing methodological frameworks while maintaining that it constitutes an independent research field ( Roche et al. 2016 ). If the field now relies heavily on existing frameworks, we question the additional value of continuing the use of “animal personality.”

Here, we ask what the recent focus on animal personality has contributed that was not already being addressed more rigorously in other fields. We address this question by asking 1) if there was ever a justified need for animal personality research to describe correlations among behaviors, 2) whether the detour into this highly descriptive field has yielded any new insights into behavioral variation, or rather if it hindered progress, and 3) whether the term “animal personality” can be safely abandoned as the field itself incorporates more rigorous and established approaches toward studying individual variation in behavior.

Individuals differ in their behavior. This is true for humans but also for nonhuman animals. When individuals differ consistently from one another, they are said to have a “personality” ( Dingemanse and Wolf 2010 ; Carere and Maestripieri 2013 ). The term “animal personality” attempts to capture 2 concepts—the first is repeatability in a behavioral response across time and context. The second concept is the notion that the expression of certain behaviors may be correlated with the expression of other behaviors, constituting a “syndrome” ( Sih et al. 2004a ). Central to the growth of the field is that the concept of animal personality has clear appeal to a nonspecialist audience—the idea that animals are like humans in that they have a “personality,” or that they may even come to have similar personalities to their owners in domesticated species ( Gosling 2001 ). Obviously, there is nothing wrong with a field of research receiving media attention, but one does need to be mindful of the possibility that the field receives instant mass-media appeal regardless of the strength of the findings or the quality of the journals in which they appear ( Dall et al. 2004 ; Pennisi 2005 ). In addition, many early personality studies were experimentally straightforward, requiring only behavioral observations and basic statistics (although these were most often incorrectly applied, see Biro and Stamps 2015 ), and were often entirely descriptive, further contributing to the ease of producing and publishing such studies. The choice of terminology was also attractive to the nonspecialist audience. “Boldness,” for example, is a much more appealing and publicly understood concept than average velocity, distance from a certain point or structure, or total distance travelled, but it is the latter behaviors that can be measured. The use of a secondary term to describe what is actually measured is therefore a marketing exercise, especially if its use adds nothing to our understanding. Moreover, because “boldness” is used to refer to different behaviors, comparisons among studies are problematic or even impossible ( David and Dall 2016 ). There may never be a universal definition of terms such as “bold,” “shy,” or “sociable,” but even if there were, it is difficult to justify their ongoing use if the only apparent value they have is increased media accessibility.

A second question is whether individual variation has indeed largely been ignored in the past. While Charles Darwin in the Origin of Species did not explicitly mention the importance of individual variation, natural selection has nothing to act on when there is no variation, as the quote at the beginning of this paper makes clear. In our view, it is therefore surprising that ignorance of this variation was held as the central justification for the invention of the new field of animal personality ( Carere and Maestripieri 2013 ). Understanding why and how behavior differs among individuals, how behavior develops over time, how selection acts on behavior, and the reasons behind the maintenance of behavioral differences are indeed all valid and important questions. Behavioral ecology as a field has never “ignored” individual variation in behavior, be it consistent across time or variable and correlated with other behaviors or not, but rather holds variation among individuals as a central component of understanding the evolution of behavior. While earlier models, aimed at understanding the origin and maintenance of individual variation, did not differentiate between among- and within-individual variation (reviewed in Dingemanse and Wolf 2010 ; Wolf and Weissing 2010 ), this does not mean understanding individual variation was not a primary and ongoing goal of the field of behavioral ecology.

There is no debate that questions about why and how behavior differs among individuals, how behavior develops over time, and how selection acts on behavior are useful and should be fundamental to understanding behavior. But we disagree with any claim that animal personality research is the only field to fully attempt to achieve this goal ( Sih et al. 2004a ). Rather, understanding variation in behavior has been the primary focus of numerous fields, such as behavioral genetics, life-history theory, alternative reproductive tactics, and foraging theory, over the past 100 years. When existing fields are relabeled, there is a danger that the rebranding creates a disconnect among the body of theory and developments that have gone before, thus giving the impression of the birth of a “new” field. In our view, such a disconnect is especially applicable to animal personality research; the field appears to have built an edifice without a foundation by largely ignoring the historical literature (see review by Crews 2013 ).

Coping styles, behavioral types, behavioral syndromes, personality, and pace-of-life syndromes have all been proposed as novel explanations for variation in behavioral phenotypes and links among behaviors ( Wilson et al. 1994 ; Gosling and John 1999 ; Sih et al. 2004b ; Nettle and Penke 2010 ; Stamps and Groothuis 2010 ; David and Dall 2016 ). These terms have a greater or lesser link to existing fields but are placed under the umbrella of “animal personality” ( David and Dall 2016 ; Roche et al. 2016 ). Such a wild growth in terminology is clearly not conducive to advancement of the field, a point made earlier by others ( Dall and Griffith 2014 ; David and Dall 2016 ; Roche et al. 2016 ), but it is not the precise definitions that are of concern, but rather whether any new insight is gained from them at all. In our view, the main advance achieved by the field of animal personality is the increased use of rigorous quantitative methods to describe variance in measures of animal behavior. However, the development of a toolkit to allow a better understanding of the statistical properties of the data collected ( Dingemanse and Wolf 2010 ) does not necessitate the use of new terminology, especially if that terminology is not used consistently across studies. While not as eye-catching as “boldness,” reporting the reaction norm intercept of a velocity measurement would also allow direct comparisons among studies. It is disappointing that despite the fact similar arguments have been made before, it remains the case that “most studies are descriptive in nature” ( Sih 2013 ) and the field focuses generally on descriptions of correlations among behavioral phenotypes and largely ignores mechanisms. The main point of progress has thus failed to translate to a shift in the actual methods employed by the bulk of the field.

While recently arguments have been made that claim that the field now does take a more hypothesis-driven approach ( Roche et al. 2016 ), a shift in animal personality that moves beyond description of behavior actually causes it to cease being “personality” research. This is because the field will then be subsumed by the approaches it borrows from a range of existing disciplines to examine behavioral plasticity and variation (see below). Once variation, consistency, and correlations among behaviors are examined in the framework of their underlying genetics, life-history, and endocrine mechanisms, these studies necessarily become part of the established fields that existed long before the term “animal personality” emerged. This, then, raises the question what the benefit of using the term has been.

A remaining point is that, despite the initial focus on behavioral consistency, personality has now come to be used to describe cases in which behavior is flexible and is not consistent over long periods of time or context (see definition in Sih 2013 ). This makes any residual definition of personality so diffuse as to be meaningless and seems to undermine the definition entirely (e.g., Galhardo et al. 2012 ). We consider this confusion to be a symptom of the paucity of hypothesis-driven research in the field. For example, at what point does consistent individual difference become behavioral plasticity , or can all behavior, no matter the degree of consistency, be called personality? Is there a statistical threshold of correlation or repeatability, below which an animal can be said not to have a personality? While Sih and Bell (2008) attempted to answer this question, the ultimate answer put forward was that there is no workable definition, and any correlation at all must be considered a behavioral type. By definition, this means formal null hypotheses cannot exist, making it impossible to reject the prior assumption that animal personality actually exists, regardless of the data collected. “Animal personality” thus defined therefore relies purely on descriptive measures, lacks explanatory power, and does not fit within the framework of an hypothesis-driven scientific method.

From the perspective of behavioral ecology, it is not clear why the above question is so frequently asked in personality research ( Muller and Chittka 2008 ; Dingemanse, Kazem, et al. 2010 ). Natural selection requires variation and behavior is subject to natural selection. Hence, the question of interest is what factors maintain variation within populations. Wolf and Weissing (2010) make the distinction between “stable and labile” states to understand why some behaviors are more variable than others. Stable states are states that are very costly, time-consuming, or even impossible to change. Behavior affected by stable states will therefore be consistent and can lead to variation at the population level but not at the individual level. Obvious stable states are those that have a strong genetic or developmental component. Students of social insects are very familiar with such states. In fact, the effect of genetics on individual behavior has been an active field of research in the social insect community for decades. In many social insects, colonies are comprised of individuals with different fathers (polyandry) or different mothers (polygyny) or sometimes both. Given the importance of high relatedness for the evolution of sociality ( Hamilton 1964a , 1964b ; Hughes et al. 2008 ), such “dilution” of relatedness requires an adaptive explanation. Many studies have shown that colonies comprised of individuals that differ in their propensity to perform certain tasks do better (reviewed in Oldroyd and Fewell 2007 ). None of these studies use the term “personality.” And if they would, the added terminology would not provide any insights into the origin of the differences. By linking the maternity and paternity of workers to the tasks they are most likely to perform, behavioral genetics, does provide us insights into the cause of the behavioral differences. Similarly, whether or not a female honeybee is raised as a queen or a worker is determined by epigenetic modification triggered by differential feeding ( Kucharski et al. 2008 ). Saying that queens and workers differ in their personality because they display behavioral differences that are consistent over time and between contexts, would be fairly considered absurd. Instead their behavioral repertoire is constrained by development, a difference of degree, but not kind, to established “personality” comparisons. The same argument could be made for any of the numerous examples of alternative reproductive tactics or phenotypes, each of which has a highly consistent suite of behaviors that would fulfill even the most strict definition of “animal personality.” When “personality” is applied to social insects, the studies revert to a simple description of consistent collective behavior with no predictive power (e.g., Wray et al. 2011 ).

In contrast to stable states, labile states can lead to behavioral variability at the individual level. Examples of labile states are gene expression levels, differences in hormone levels or energy reserves. Because in order to qualify as a “personality,” behavior needs to be consistent (but over what period is not defined [ Sih 2013 ]), it seems that labile states pose a problem. Labile states thus require an explanation so that their presence does not undermine the concept of “personality.” Hormone and energy levels in particular can fluctuate wildly over relatively short time spans, so how can they result in consistent behavior? Wolf and Weissing (2010) argue that labile states can become nonlabile via positive feedback mechanisms. If a hungry individual continues to be successful in obtaining food, its energy levels will remain high. When energy levels affect foraging behavior, for instance how likely an organism is to take risks while foraging, then the argument goes that positive feedback results in behavioral consistency. Similar arguments have been made in the past and are known collectively as optimal foraging theory ( Charnov 1976 ; Kacelnik and Bateson 1996 ; Perry and Pianka 1997 ). More generally, state-dependent behavior is a concept that was developed in the 1980s to understand behavioral variation and how selection acts on such variation ( Houston and Davies 1985 ; Houston and McNamara 1988 ). State-dependent behavior can be subject to frequency dependence when a particular behavioral repertoire affects that of an alternative behavioral repertoire ( Wolf and Weissing 2010 ), an insight that has been around since Maynard-Smith’s pivotal work on game theory and evolutionary stable strategies ( Maynard Smith 1982 ).

Dingemanse, Kazem, et al. (2010) in fact argue that it is an organism’s behavioral response over an environmental gradient (“context”) that is the trait of interest. They therefore use the term behavioral reaction norm to describe the set of behavioral phenotypes that a single individual produces in a given set of environments. When an individual behaves differently in different environments, in other words its reaction norm is nonhorizontal, it shows phenotypic plasticity. The authors themselves directly link behavioral reaction norms to phenotypic plasticity, an important driving force in adaptive evolution. If this is taken as a novel insight, it is worth comparing it to the origin of the concept of phenotypic plasticity. In 1896, Baldwin argued for a “new factor in evolution.” This “new” factor referred to traits acquired during the ontogeny of the organism. Baldwin called this new factor “the organic selection” factor. Nowadays, we would refer to such factor as phenotypic plasticity, and there is a great deal of research on and understanding about behavioral plasticity, for example in the context of niche specialization ( Bolnick et al. 2003 ). Even though the inclusion of reaction norms into personality research has encouraged a stronger statistical basis, the introduction of a new term itself—behavioral reaction norm—has not added new insight into the evolution of behavioral variation.

Much of the emphasis for animal personality studies comes from a purported increase in our ability to understand how selection acts on suites of behaviors. Because the behavioral phenotype itself is not carried across generations—only the mechanism that generates that behavior, be it genetic, epigenetic, or even cultural—understanding selection on correlated phenotypes without insight into a mechanism that may respond to selection is a cursory approach. A central claim made by personality researchers is that they take a more integrative approach to understanding why animals behave the way they do ( Réale et al. 2010 ), and that “animal personality” was the first field to assess the role of hormones, physiology, and metabolism in shaping an organism’s behavior ( Réale et al. 2010 ). Yet, Wolf and Weissing (2010) in the same issue explicitly mention that behavioral correlations can often be understood in terms of the genetic, physiological, neurobiological and cognitive systems that underlie behavior, and proceed to mention many examples from the literature. In fact, life-history evolution had already incorporated behavior, genetics, physiology, neurobiology, and cognition ( Roff 1992 ; Stearns 1992 ). The whole point of life-history theory is to understand how natural selection acts on physiology, anatomy, and behavior to affect behavior and life span. This is true regardless of whether the behaviors studied are correlated with others or the time scale over which they vary. Moreover, a crucial aspect of life-history theory is that of trade-offs—the notion that behaviors may be selected for strongly in one domain to the detriment of performance in another. The idea of trade-offs has permeated the personality literature to explain why a particular behavior or suite of behaviors may be adaptive in one context but not the other ( Sih et al. 2004a ; Dingemanse, Kazem, et al. 2010 ), but this cannot be considered an insight generated by personality research, but rather as another example of conceptual appropriation and relabeling by animal personality researchers.

Definitions of behavioral syndromes, such as an overall “bold” phenotype, are based on measuring correlations among a suite of behaviors, frequently using measurements like emergence time, exploration, time out from cover, distance from a novel object, and speed of movement. That these behaviors are correlated within individuals is often taken as a meaningful signature of selection or a common underlying mechanism, but a necessary null hypothesis that they are in fact different interpretations of the same measurement must surely be considered. For example, individual speed functions as an explanatory factor for numerous aspects of social interactions and movement that might otherwise be considered as separate social behaviors ( Katz et al. 2011 ; Tunstrøm et al. 2013 ). Therefore, a “behavioral syndrome” may actually be composed of the same behavior measured and interpreted in different ways, and therefore based firmly in life-history theory and biophysics rather than animal personality. This is generally captured in Morgan’s Canon that “In no case is an animal activity to be interpreted in terms of higher psychological processes if it can be fairly interpreted in terms of processes which stand lower in the scale of psychological evolution and development” ( Morgan 1903 ). Approaches such as those employed in computational ethology and unsupervised machine learning ( Kabra et al. 2013 ; Anderson and Perona 2014 ) further remove the need for a priori labels and, with their increased usage, will render arbitrary divisions among behaviors or groups of behaviors irrelevant.

One physiological trait that likely affects activity levels, growth, and fecundity is an animal’s resting metabolic rate ( Hulbert et al. 2007 ). Thus, animals with high resting metabolic rates can be classified as “bold” individuals ( Biro and Stamps 2010 ), simply because resting metabolic rate influences a suite of behaviors. More generally, we can see energy metabolism as a general physiological explanation for individual differences simply because levels of energy affect almost anything an organism does or can do ( Biro and Stamps 2010 ). An animal with a higher metabolism moves faster and therefore explores greater areas and is more likely to enter new environments ( Réale et al. 2010 ). It therefore offers no insight to conclude that ectotherms become more “bold” as temperature increases due to an increase in their activity levels ( Forsatkar et al. 2016 ). Likewise, individuals with higher levels of testosterone are more aggressive, potentially allowing them to compete successfully with conspecifics over access to both food or mates, although carrying costs in the form of reduced immune function or increased predation risk. And indeed, Réale et al. (2010) argue that the correlation between metabolism and “personality” (and hormones and “personality”) is an example of a “pace-of-life” syndrome, which, in turn, really refers to a trade-off between longevity and metabolic rate ( Hulbert et al. 2007 ).

Consider a population of white-throated sparrows, Zonotrichia albicollis . Male and female sparrows come in 2 morphs: tan or white, based on the color of the median crown stripe. White males are more aggressive, frequently intrude into neighboring territories, spend less time guarding their mates, occasionally attempt polygyny, and provide less parental care than tan males ( Tuttle 2003 ). It is clear that a male cannot simultaneously pursue additional matings and provide high levels of paternal care to his young. Thus, white and tan morph males practice different reproductive strategies based on the trade-off between securing additional matings and parental effort. In order to be successful in obtaining extrapair copulations, white males behave aggressively and frequently intrude into another male’s territory to gain access to mates. We could thus argue that white males are bold, whereas tan males are shy. But does that reveal the underlying cause of the behavioral differences between white and tan males? It does not. What we know is that white and tan males differ in a suite of behaviors.

We cannot understand the white-throated sparrow males’ behavior if we ignore the behavior of the females. White-throated sparrows mate disassortatively so that most white males pair up with tan females and vice versa. When a tan female’s mate reduces his parental effort because he is pursuing extrapair copulations, the tan female will compensate by increasing her parental effort. This basically “allows” the white males to increase his reproductive output via extrapair copulations as his earlier brood’s future is secured thanks to the female’s parental effort. Tan males are not afforded such luxury, as their white mates will not compensate for a reduction in parental effort, thus forcing tan males to remain faithful. Hence, both reproductive strategies are maintained in the population because both color morphs have equal reproductive success ( Tuttle 2003 ). Genetically polymorphism is maintained because of “supergenes”, linked clusters of coevolved genes that give rise to divergent fitness-related traits that are variable within species ( Schwander et al. 2014 ; Tuttle et al. 2016 ). Hence, while personality researchers would have been satisfied with identifying behavioral differences (e.g., Malange et al. 2016 ; Michelangeli et al. 2016 ; Yuen et al. 2016 ), the fields of life-history evolution and behavioral genetics allows us to understand why individuals differ in their behavior. This is just one of many examples of studies that have identified the genetic mechanisms underlying animal behavior. Whether or not a worker honeybee becomes reproductively active is determined by a single gene, called anarchy (Ronai, Oldroyd, et al. 2016 ; Ronai, Vergoz, et al., 2016 ) and the social structure of several species of ants is determined by a “social chromosome” ( Purcell et al. 2014 ). Similarly, the kind of burrow dug by deer mice is under the influence of a set of well-characterized genes ( Hu and Hoekstra 2017 ). Identifying the genetic basis of behavior is obviously much more labor intensive, not to mention expensive, than simply describing differences in “personality,” but actually provides us insights into the underlying mechanisms.

Another well-understood mechanism mediating suites of behavior can be found in the African cichlid Astatotilapia burtoni. Males of this species display distinct behavioral phenotypes, depending on their social context (cf. Bergmüller and Taborsky 2010 ) and condition. Males form social hierarchies in which a small number of males are highly aggressive, territorial, brightly colored, swim rapidly, have large ranges, and are reproductively active. The other males in these groups are submissive, have reduced coloration, swim slowly, and have a restricted range ( Maruska 2015 ). These differences in behavior and physiology are based on the relative social position and relationships to other individuals in the social network. Males can rapidly transition among these states based on social context, leading to massive changes in individual behaviors, as well as the correlations among these behaviors. These social phenotypes are plastic and reversible, meaning that individual males may frequently switch between dominant and subordinate social status.

Probably because the physiological mechanisms underlying the transition between the above described behavioral states, and the links among behaviors are well characterized, there has been no need or benefit to describe them as personality differences. When the social environment changes such that males either ascend (subordinate to dominant) or descend (dominant to subordinate) in rank, there are rapid changes in behavior, circulating hormones, and levels of gene expression in the brain that reflect the direction of transition ( Maruska 2014 ). Socially ascendant and dominant males have increased activation of brain nuclei in the social behavior network, and higher levels of sex steroids in the plasma, which affect numerous behaviors. Ascending males also show rapid changes in levels of neuropeptide and steroid receptors in the brain, as well as in the pituitary and testes (reviewed in Maruska 2014 ). Although these dominance changes result in massive and correlated behavioral changes, which are consistent within individuals and over time, the deeper understanding of the mechanisms underlying this switch seem to preclude the need to refer to these behavioral states as “personality.” This example again emphasizes how little is gained from labels such as “bold” and “shy” in the face of rigorous inquiry into the causes and consequences of fixed and labile behaviors.

There is nothing inherently wrong with constructing a new research field around animal behavior that leans heavily on existing fields, particularly if the claimed aim is to integrate separate research areas. After all the field of behavioral ecology has its origin in the integration of ethology, evolutionary biology, game theory, and economic decision-making. While ethology traditionally described how animals behave, behavioral ecology started to ask why they behave in a particular way and not another way. The incorporation of economic thinking and game theory into the study of animal behavior increased the biologist’s mathematical tool kit, allowing researchers to construct models that predict how organisms should behave under particular conditions. Combined with the realization that selection acts at the level of genes and not groups or individuals (with some exceptions) has allowed behavioral ecology to achieve what ethology cannot: understand how selection shapes the behavior of animals in different contexts.

Behavioral ecology can explain why individuals sometimes appear to behave against their own interest. Why abandon your young so that they are bound to die? One answer is that if a parent is forced to look after young alone because the other parent has left first, it may be in the remaining parent’s best interest to secure resources for the next breeding season instead of attempting to raise young alone. Some parents may be in sufficient condition to be able to successfully raise their young alone, particularly when environmental conditions too are favorable, whereas a parent in poor condition is forced to call it quits. Thus, both differences in physiology and environmental conditions can cause behavioral differences among individuals faced with the same choice. In the field of animal personality, one might claim that the 2 individuals have different personalities, even as a function of their life-history, but what does that tell us we did not know already without reference to personality? Herein lies the crux of the matter. For a research field to be of value, it needs to be hypothesis driven and have predictive power. Simply coining individual differences “personalities” has no predictive power. To be convinced that animal personality research should be taken seriously, we need clear hypotheses and predictions. In the absence of both, we equate “animal personality” with “ethology”; a well-established field that describes behavior. Indeed, animals behave and they do not all behave in the same way. But in order to advance the field, we need to understand why animals differ in the way they behave. The expanding use of quantitative methods to study behavioral variation is an advance but has diverged from and should no longer be considered part of personality research, instead becoming a conduit between the existing fields of behavioral ecology and quantitative genetics.

A general and pervasive problem in animal personality research is that the problems and questions it claims to address were, and continue to be addressed more rigorously in existing fields. In our view the detour into, and then out of, personality obfuscates the history and development of our scientific understanding of behavioral variation and plasticity. This can present a serious barrier to understanding for new students interested in this field, because they may be left with the impression that variation in behavior can be understood using simple correlational approaches to behavioral phenotypes that typify the bulk of the field. We therefore hope that our piece will initiate an honest debate regarding the validity of animal personality research, while highlighting the more valuable offshoots that diverged in the early days of the field. If indeed the use of “animal personality” does not add anything new to behavioral ecology, as we argue, then let us please resume discussion of variation and consistency of behavior in the context of behavioral ecology, and abandon the concept and terminology of animal personality entirely.

M.B. is supported by the Australian Research Council (FT120100120). L.A.J. is supported by the Minerva Foundation grant 712432.

Anderson DJ Perona P . 2014 . Toward a science of computational ethology . Neuron . 84 : 18 – 31 .

Google Scholar

Araya-Ajoy YG Mathot KJ Dingemanse NJ . 2015 . An approach to estimate short-term, long-term and reaction norm repeatability . Methods Ecol Evol . 6 : 1462 – 1473 .

Baldwin JM . 1896 . A new factor in evolution . Am Nat . 30 : 441 – 451, 536–553 .

Bell AM Hankison SJ Laskowski KL . 2009 . The repeatability of behaviour: a meta-analysis . Anim Behav . 77 : 771 – 783 .

Bergmüller R Taborsky M . 2010 . Animal personality due to social niche specialisation . Trends Ecol Evol . 25 : 504 – 511 .

Biro PA Stamps JA . 2010 . Do consistent individual differences in metabolic rate promote consistent individual differences in behavior? Trends Ecol Evol . 25 : 653 – 659 .

Biro PA Stamps JA . 2015 . Using repeatability to study physiological and behavioural traits . Anim Behav . 105 : 223 – 230 .

Bolnick DI Svanbäck R Fordyce JA Yang LH Davis JM Hulsey CD Forister ML . 2003 . The ecology of individuals: incidence and implications of individual specialization . Am Nat . 161 : 1 – 28 .

Carere C Maestripieri D , editors. 2013 . Animal personalities. Behavior, physiology, and evolution . Chicago (IL) : The University of Chicago Press .

Google Preview

Charnov EL . 1976 . Optimal foraging, the marginal value theorem . Theor Popul Biol . 9 : 129 – 136 .

Crews D . 2013 . Animal personalities: behavior, physiology, and evolution . Claudio Carere and Dario Maestripieri , editors. Integr Comp Biol . 53 : 873 – 875 .

Dall SRX Griffith SC . 2014 . An empiricist guide to animal personality variation in ecology and evolution . Frontiers Ecol Evol . doi: 10.3389/fevo.2014.00003 .

Dall SRX Houston AI McNamara JM . 2004 . The behavioural ecology of personality: consistent individual differences from an adaptive perspective . Ecol Lett . 7 : 734 – 739 .

David M Dall SRX . 2016 . Unravelling the philosophies underlying ‘animal personality’ studies: a brief re-appraisal of the field . Ethology . 122 : 1 – 9 .

Dingemanse NJ Dochtermann N Wright J . 2010a . A method for exploring the structure of behavioural syndromes to allow formal comparison within and between data sets . Anim Behav . 79 : 439 – 450 .

Dingemanse NJ Dochtermann NA . 2014 . Individual behaviour: behavioural ecology meets quantitative genetics . In: Charmantier A Garant D Kruuk LEB , editors. Quantitative genetics in the wild . Oxford : Oxford University Press . p. 54 – 67 .

Dingemanse NJ Kazem AJ Réale D Wright J . 2010 . Behavioural reaction norms: animal personality meets individual plasticity . Trends Ecol Evol . 25 : 81 – 89 .

Dingemanse NJ Wolf M . 2010 . Recent models for adaptive personality differences: a review . Philos Trans R Soc Lond B Biol Sci . 365 : 3947 – 3958 .

Forsatkar MN Nematollahi MA Biro PA Beckmann C . 2016 . Individual boldness traits influenced by temperature in male Siamese fighting fish . Physiol Behav . 165 : 267 – 272 .

Galhardo L Vitorino A Oliveira RF . 2012 . Social familiarity modulates personality trait in a cichlid fish . Biol Lett . 8 : 936 – 938 .

Gosling SD . 2001 . From mice to men: what can we learn about personality from animal research? Psychol Bull . 127 : 45 – 86 .

Gosling SD John OP . 1999 . Personality dimensions in nonhuman animals: a cross-species review . Curr D Psych Sc . 8 : 69 – 75 .

Hamilton WD . 1964a . The genetical evolution of social behaviour. I . J Theor Biol . 7 : 1 – 16 .

Hamilton WD . 1964b . The genetical evolution of social behaviour. II . J Theor Biol . 7 : 17 – 52 .

Houston AI Davies NB . 1985 . The evolution of cooperation and life history in the dunnock, Prunella modularis . In: Sibly RM Smith RH , eds. Behavioural ecology ecological consequences of adaptive behaviour . Oxford : Blackwell Scientific Publications . p. 471 – 487 .

Houston AI McNamara JN . 1988 . The ideal free distribution when competitive abilities differ: an approach based on statistical mechanics . Anim Behav . 36 : 166 – 174 .

Hu CK Hoekstra HE . 2017 . Peromyscus burrowing: a model system for behavioral evolution . Semin Cell Dev Biol . 61 : 107 – 114 .

Hughes WO Oldroyd BP Beekman M Ratnieks FL . 2008 . Ancestral monogamy shows kin selection is key to the evolution of eusociality . Science . 320 : 1213 – 1216 .

Hulbert AJ Pamplona R Buffenstein R Buttemer WA . 2007 . Life and death: metabolic rate, membrane composition, and life span of animals . Physiol Rev . 87 : 1175 – 1213 .

Kabra M Robie AA Rivera-Alba M Branson S Branson K . 2013 . JAABA: interactive machine learning for automatic annotation of animal behavior . Nat Methods . 10 : 64 – 67 .

Kacelnik A Bateson M . 1996 . Risky theories—the effects of variance on foraging decisions . Amer Zool . 36 : 402 – 434 .

Katz Y Tunstrøm K Ioannou CC Huepe C Couzin ID . 2011 . Inferring the structure and dynamics of interactions in schooling fish . Proc Nat Acad Sci . 108 : 18720 – 18725 .

Kucharski R Maleszka J Foret S Maleszka R . 2008 . Nutritional control of reproductive status in honeybees via DNA methylation . Science . 319 : 1827 – 1830 .

Malange J Izar P Japyassú H . 2016 . Personality and behavioural syndrome in Necromys lasiurus (Rodentia: Cricetidae): notes on dispersal and invasion processes . Acta Ethol . 19 : 189 – 195 .

Maruska KP . 2014 . Social regulation of reproduction in male cichlid fishes . Gen Comp Endocrinol . 207 : 2 – 12 .

Maruska KP . 2015 . Social transitions cause rapid behavioral and neuroendocrine changes . Integr Comp Biol . 55 : 294 – 306 .

Mathot KJ Wright J Kempenaers B Dingemanse NJ . 2012 . Adaptive strategies for managing uncertainty may explain personality-related differences in behavioural plasticity . Oikos . 121 : 1009 – 1020 .

Maynard Smith J . 1982 . Evolution and the theory of games . Cambridge (UK) : Cambridge University Press .

McElreath R Strimling P . 2006 . How noisy information and individual asymmetries can make ‘personality’ an adaptation: a simple model . Anim Behav . 72 : 1135 – 1139 .

Michelangeli M Chapple DG Wong BBM . 2016 . Are behavioural syndromes sex specific? Personality in a widespread lizard species . Behav Ecol & Sociobiol . 70 : 1911 – 1919 .

Morgan CL . 1903 . An introduction to comparative psychology . London : W. Scott .

Muller H Chittka L . 2008 . Animal personalities: the advantage of diversity . Curr Biol . 18 : R961 – R963 .

Nettle D Penke L . 2010 . Personality: bridging the literatures from human psychology and behavioural ecology . Philos Trans R Soc Lond B Biol Sci . 365 : 4043 – 4050 .

Oldroyd BP Fewell JH . 2007 . Genetic diversity promotes homeostasis in insect colonies . Trends Ecol Evol . 22 : 408 – 413 .

Pennisi E . 2005 . Animal behavior. Strong personalities can pose problems in the mating game . Science . 309 : 694 – 695 .

Perry G Pianka ER . 1997 . Animal foraging: past, present and future . Trends Ecol Evol . 12 : 360 – 364 .

Purcell J Brelsford A Wurm Y Perrin N Chapuisat M . 2014 . Convergent genetic architecture underlies social organization in ants . Curr Biol . 24 : 2728 – 2732 .

Réale D Garant D Humphries MM Bergeron P Careau V Montiglio PO . 2010 . Personality and the emergence of the pace-of-life syndrome concept at the population level . Philos Trans R Soc Lond B Biol Sci . 365 : 4051 – 4063 .

Roche DG Careau V Binning SA . 2016 . Demystifying animal ‘personality’ (or not): why individual variation matters to experimental biologists . J Exp Biol . 219 : 3832 – 3843 .

Roff DA . 1992 . The evolution of life histories. Theory and analysis . New York : Chapman & Hall .

Ronai I Oldroyd BP Barton DA Cabanes G Lim J Vergoz V . 2016 . Anarchy is a molecular signature of worker sterility in the honey bee . Mol Biol Evol . 33 : 134 – 142 .

Ronai I Vergoz V Oldroyd BP . 2016 . The mechanistic, genetic, and evolutionary basis of worker sterility in the social Hymenoptera . Adv Study Behav . 48 : 251 – 317 .

Schwander T Libbrecht R Keller L . 2014 . Supergenes and complex phenotypes . Curr Biol . 24 : R288 – R294 .

Sih A . 2013 . Frontiers on the interface between behavioral syndromes and social behavioral ecology . In: Carere C Maestripieri D , editors. Animal personalities: behavior, physiology, and evolution . Chicago (IL) : Chicago University Press . p. 221 – 251 .

Sih A Bell A Johnson JC . 2004a . Behavioral syndromes: an ecological and evolutionary overview . Trends Ecol Evol . 19 : 372 – 378 .

Sih A Bell AM . 2008 . Insights for behavioral ecology from behavioral syndromes . Adv Study Behav . 38 : 227 – 281 .

Sih A Bell AM Johnson JC Ziemba RE . 2004b . Behavioral syndromes: an integrative overview . Q Rev Biol . 79 : 241 – 277 .

Stamps JA Groothuis TG . 2010 . Developmental perspectives on personality: implications for ecological and evolutionary studies of individual differences . Philos Trans R Soc Lond B Biol Sci . 365 : 4029 – 4041 .

Stearns SC . 1992 . The evolution of life histories . Oxford : Oxford University Press .

Tunstrøm K Katz Y Ioannou CC Huepe C Lutz MJ Couzin ID . 2013 . Collective states, multistability and transitional behavior in schooling fish . PLoS Comput Biol . 9 : e1002915 .

Tuttle EM . 2003 . Alternative reproductive strategies in the white-throated sparrow: behavioral and genetic evidence . Behav Ecol . 14 : 425 – 432 .

Tuttle EM Bergland AO Korody ML Brewer MS Newhouse DJ Minx P Stager M Betuel A Cheviron ZA Warren WC et al.  . 2016 . Divergence and functional degradation of a sex chromosome-like supergene . Curr Biol . 26 : 344 – 350 .

Westneat DF Wright J Dingemanse NJ . 2015 . The biology hidden inside residual within-individual phenotypic variation . Biol Rev Camb Philos Soc . 90 : 729 – 743 .

Wilson DS Clark AB Coleman K Dearstyne T . 1994 . Shyness and boldness in humans and other animals . Trends Ecol Evol . 9 : 442 – 445 .

Wolf M McNamara JM . 2012 . On the evolution of personalities via frequency-dependent selection . Am Nat . 179 : 679 – 692 .

Wolf M van Doorn GS Leimar O Weissing FJ . 2007 . Life-history trade-offs favour the evolution of animal personalities . Nature . 447 : 581 – 584 .

Wolf M van Doorn GS Weissing FJ . 2008 . Evolutionary emergence of responsive and unresponsive personalities . Proc Nat Acad Sci . 105 : 15825 – 15830 .

Wolf M Van Doorn GS Weissing FJ . 2011 . On the coevolution of social responsiveness and behavioural consistency . Proc Roy Soc B: Biol Sci . 278 : 440 – 448 .

Wolf M Weissing FJ . 2010 . An explanatory framework for adaptive personality differences . Philos Trans R Soc Lond B Biol Sci . 365 : 3959 – 3968 .

Wray MK Mattila HR Seeley TD . 2011 . Collective personalities in honeybee colonies are linked to colony fitness . Anim Behav . 81 : 559 – 568 .

Yuen CH Pillay N Heinrichs M Schoepf I Schradin C . 2016 . Personality traits are consistent when measured in the field and in the laboratory in African striped mice ( Rhabdomys pumilio ) . Behav Ecol & Sociobiol . 70 : 1235 – 1246 .

Author notes

Email alerts

Citing articles via.

• Recommend to your Library

Affiliations

• Online ISSN 1465-7279
• Copyright © 2024 International Society of Behavioral Ecology
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Transforming the understanding and treatment of mental illnesses.

Información en español

Celebrating 75 Years! Learn More >>

• About the Acting NIMH Director
• Advisory Boards and Groups
• Strategic Plan
• Offices and Divisions
• Careers at NIMH
• Staff Directories
• Getting to NIMH

A Hypothesis-Based Approach: The Use of Animals in Mental Health Research

By Joshua Gordon

October 21, 2019

This Director’s Message on the use of animal models in mental health research is one of two focused on this topic and is aimed at the research community. A companion message, written for a general audience, is titled: What Can Animals Tell Us About Mental Illnesses?

Model organisms play a crucial role in modern neuroscience. Exploring the function of molecules, cells, circuits, and systems and how they relate to behavior often requires the use of methods to examine the intact brain that, for ethical and practical reasons, can only be performed in animals. Yet, the translation of results gained from animals into treatments for humans has been challenging. At the National Institute of Mental Health (NIMH), we have been engaged in a year-long interactive process to understand and clarify the appropriate role of animal research in our portfolio. While we continue to refine our policies, this message and an accompanying notice  are meant to communicate our current policies regarding the scientifically justified use of animals in mental health research.

Utility rather than validity: A hypothesis-based approach

First and foremost, we must realize that there is no such thing as a true animal model of a psychiatric disorder. Such models are, at best, approximations. How then should we evaluate animal models? For the past few years, we at NIMH have been asking the following questions about studies involving model organisms:

What is the question being asked?

Is it an important question.

• Does the proposed experimental system enable that question to be answered?

These questions support an evaluation of the appropriateness and impact of a hypothesis-based approach to studying the biology of mental illnesses using animal models.

Researchers create models to test hypotheses. We believe that NIMH-supported studies should not be aimed at establishing whether a particular model has validity for understanding a human mental illness. Instead, they should be aimed at testing a specific hypothesis. Too frequently, we receive proposals in which an investigator proposes to create an animal model of a disorder and “validate” that animal model by running a series of assays to document similarities between the animal and humans with the disorder. Such studies are problematic for many reasons, not the least of which is low statistical rigor, given that positive publication bias coupled with multiple tests can often lead investigators (and those that follow) astray. A strong focus on a mechanistic hypothesis, with the animal model designed to test that hypothesis, coupled with a rigorously planned and sufficiently powered experimental design, increases the reproducibility of the results. Moreover, focusing on the question ensures that the knowledge gained can be built upon by succeeding investigators.

For the purposes of grant applications, we urge that investigators explicitly set forth the hypothesis upfront and explain how the proposed experiments address this hypothesis. Notably, the hypothesis can be driven by basic science interests. How does disrupting a given cellular function alter the development of a neural circuit? What is the role of that circuit element in a particular behavior? How do multiple brain regions interact during that behavior? They may also be driven by clinical questions, testing hypotheses that arise from studies in patients. Do disruptions in inhibitory interneurons alter prefrontal function? What is the role of sensory inputs to the amygdala in social recognition of emotion? How does exposure to adverse environments during development affect cortico-limbic interactions? Each of these questions addresses fundamental areas of biology of relevance to mental illnesses.

There are some areas of neurobiology where we have identified specific priority areas and additional topics where we might require detailed justification given known challenges. Several of these areas are delineated in the accompanying  notice  . For example, fundamental studies of the biology of genes and gene products implicated in mental illnesses by unbiased approaches using genome-wide significance thresholds are of particular interest to NIMH given their definitive relationship to the human conditions. Conversely, genes previously identified via candidate approaches and not subsequently verified by genome-wide approaches are of considerably less interest. These and other priorities in the area of genetics and genomics are clearly stated in guidance we have published following the Report of the National Advisory Mental Health Council Genomics Workgroup .

Similarly, we are prioritizing computational behavioral phenotypes over simplistic, pharmacologically-validated behavioral tests for reasons to do with specificity and clarity of mechanism. For example, traditional behavioral responses to stress paradigms are particularly problematic. Non-specific tests such as the forced-swim or tail suspension tests, among others, have largely failed to reveal translatable neural mechanisms, and lack specificity from a pharmacologic-validity perspective. Approaches that examine and rigorously quantify the impacts of stress on reward and arousal systems, by contrast, are promising avenues of research that hold considerable promise to reveal novel mechanistic insights and lead to new therapeutic avenues.

Does the experimental system proposed enable that question to be answered?

The issue of the usefulness of an animal model cannot be divorced from the question that the model is trying to answer. Precision here is crucial, as are the measured variables. For example, consider facial expressions of emotion. Mice don’t have them. But mice do have measurable facial expressions of physical pain. If an investigator wants to understand the neural pathways leading to facial expressions of emotion, mouse models will not help. But if the goal is to understand the pathways leading from pain to facial motor output and how they are modulated during behavior, the mouse might be useful.

The point here is that one should ask the question first and then figure out what the best experimental system is to answer that question. The evaluation of the experimental system should include consideration of the ethical and efficient use of resources, the feasibility of the approach, and the potential evolutionary conservation of the mechanism of interest. Most questions will probably require a variety of systems in order to answer them fully, maximize rigor and reproducibility, and ensure translatability. For many investigators, this will mean collaborations and team science. But for investigators using animals, this means clarifying why the particular model organism was chosen and how that model facilitates hypothesis testing.

We are here to help

We have attempted to make the NIMH position on animal models as clear as possible, but there are a number of gray areas, and every investigator’s situation is unique. In addition to examining the resources we have available on the web — including this message, the notice, and other materials — I encourage as always communication with NIMH program staff early in the grant preparation process to ensure that potential issues are addressed appropriately in advance.

• History & Society
• Science & Tech
• Biographies
• Animals & Nature
• Geography & Travel
• Arts & Culture
• Games & Quizzes
• On This Day
• One Good Fact
• New Articles
• Lifestyles & Social Issues
• Philosophy & Religion
• Politics, Law & Government
• World History
• Health & Medicine
• Browse Biographies
• Birds, Reptiles & Other Vertebrates
• Bugs, Mollusks & Other Invertebrates
• Environment
• Fossils & Geologic Time
• Entertainment & Pop Culture
• Sports & Recreation
• Visual Arts
• Demystified
• Image Galleries
• Infographics
• Top Questions
• Britannica Kids
• Saving Earth
• Space Next 50
• Student Center
• Introduction

Darwin’s influence

Ecological and ethological approaches to the study of behaviour.

• Sensory-motor mechanisms
• Cognitive mechanisms
• Behavioral genetics
• Instinctive learning
• Natural selection in action
• Adaptive design
• Character mapping
• Phylogenetic grading
• Artificial selection
• The comparative approach

History and basic concepts

Our editors will review what you’ve submitted and determine whether to revise the article.

• Khan Academy - Intro to animal behavior
• Psychology Today - Animal Behavior
• Biology LibreTexts Library - Evolution of Animal Behavior
• National Center for Biotechnology Information - PubMed Central - The Science of Animal Behavior and Welfare: Challenges, Opportunities, and Global Perspective
• animal behavior - Children's Encyclopedia (Ages 8-11)
• animal behavior - Student Encyclopedia (Ages 11 and up)
• Table Of Contents

The origins of the scientific study of animal behaviour lie in the works of various European thinkers of the 17th to 19th centuries, such as British naturalists John Ray and Charles Darwin and French naturalist Charles LeRoy. These individuals appreciated the complexity and apparent purposefulness of the actions of animals, and they knew that understanding behaviour demands long-term observations of animals in their natural settings. At first, the principal attraction of natural history studies was to confirm the ingenuity of God. The publication of Darwin’s On the Origin of Species in 1859 changed this attitude. In his chapter on instinct , Darwin was concerned with whether behavioral traits, like anatomical ones, can evolve as a result of natural selection . Since then, biologists have recognized that the behaviours of animals, like their anatomical structures, are adaptations that exist because they have, over evolutionary time (that is, throughout the formation of new species and the evolution of their special characteristics), helped their bearers to survive and reproduce.

Furthermore, humans have long appreciated how beautifully and intricately the behaviours of animals are adapted to their surroundings. For example, young birds that possess camouflaged colour patterns for protection against predators will freeze when the parent spots a predator and calls the alarm. Darwin’s achievement was to explain how such wondrously adapted creatures could arise from a process other than special creation. He showed that adaptation is an inexorable result of four basic characteristics of living organisms:

• There is variation among individuals of the same species . Even closely related individuals, such as parent and offspring or sibling and sibling, differ considerably. Familiar human examples include differences in facial features, hair and eye colour , height, and weight.
• Many of these variations are inheritable —that is, offspring resemble their parents in many traits as a result of the genes they share.
• There are differences in numbers of surviving offspring among parents in every species. For example, one female snapping turtle (family Chelydridae) may lay 24 eggs ; however, only 5 may survive to adulthood. In contrast, another female may lay only 18 eggs, with 1 of her offspring surviving to adulthood.
• The individuals that are best equipped to survive and reproduce perpetuate the highest frequency of genes to descendant populations. This is the principle known colloquially as “ survival of the fittest ,” where fitness denotes an individual’s overall ability to pass copies of his genes on to successive generations. For example, a woman who rears six healthy offspring has greater fitness than one who rears just two.

An inevitable consequence of variation, inheritance , and differential reproduction is that, over time, the frequency of traits that render individuals better able to survive and reproduce in their present environment increases. As a result, descendant generations in a population resemble most closely the members of ancestral populations that were able to reproduce most effectively. This is the process of natural selection.

The natural history approach of Darwin and his predecessors gradually evolved into the twin sciences of animal ecology , the study of the interactions between an animal and its environment , and ethology , the biological study of animal behaviour. The roots of ethology can be traced to the late 19th and early 20th centuries, when scientists from several countries began exploring the behaviours of selected vertebrate species: dogs by the Russian physiologist Ivan Pavlov ; rodents by American psychologists John B. Watson , Edward Tolman , and Karl Lashley ; birds by American psychologist B.F. Skinner ; and primates by German American psychologist Wolfgang Köhler and American psychologist Robert Yerkes . The studies were carried out in laboratories, in the case of dogs, rodents and pigeons , or in artificial colonies and laboratories, in the case of primates. These studies were oriented toward psychological and physiological questions rather than ecological or evolutionary ones.

It was not until the 1930s that field naturalists—such as English biologist Julian Huxley , Austrian zoologist Konrad Lorenz , and Dutch-born British zoologist and ethologist Nikolaas Tinbergen studying birds and Austrian zoologist Karl von Frisch and American entomologist William Morton Wheeler examining insects —gained prominence and returned to broadly biological studies of animal behaviour. These individuals, the founders of ethology , had direct experience with the richness of the behavioral repertoires of animals living in their natural surroundings. Their “return to nature” approach was, to a large extent, a reaction against the tendency prevalent among psychologists to study just a few behavioral phenomena observed in a handful of species that were kept in impoverished laboratory environments .

The goal of the psychologists was to formulate behavioral hypotheses that claimed to have general applications (e.g., about learning as a single, all-purpose phenomenon). Later they would proceed using a deductive approach by testing their hypotheses through experimentation on captive animals. In contrast, the ethologists advocated an inductive approach , one that begins with observing and describing what animals do and then proceeds to address a general question: Why do these animals behave as they do? By this they meant “How do the specific behaviours of these animals lead to differential reproduction?” Since its birth in the 1930s, the ethological approach—which stresses the direct observation of a broad array of animal species in nature, embraces the vast variety of behaviours found in the animal kingdom, and commits to investigating behaviour from a broad biological perspective—has proved highly effective.

One of Tinbergen’s most important contributions to the study of animal behaviour was to stress that ethology is like any other branch of biology , in that a comprehensive study of any behaviour must address four categories of questions, which today are called “levels of analysis,” including causation , ontogeny , function , and evolutionary history . Although each of these four approaches requires a different kind of scientific investigation, all contribute to solving the enduring puzzle of how and why animals, including humans, behave as they do. A familiar example of animal behaviour—a dog wagging its tail—serves to illustrate the levels of analysis framework. When a dog senses the approach of a companion (dog or human), it stands still, fixates on the approaching individual, raises its tail, and begins swishing it from side to side. Why does this dog wag its tail? To answer this general question, four specific questions must be addressed.

With respect to causation, the question becomes: What makes the behaviour happen? To answer this question, it becomes important to identify the physiological and cognitive mechanisms that underlie the tail-wagging behaviour. For example, the way the dog’s hormonal system adjusts its responsiveness to stimuli, how the dog’s nervous system transmits signals from its brain to its tail, and how the dog’s skeletal-muscular system generates tail movements need to be understood. Causation can also be addressed from the perspective of cognitive processes (that is, knowing how the dog processes information when greeting a companion with tail wagging). This perspective includes determining how the dog senses the approach of another individual, how it recognizes that individual as a friend, and how it decides to wag its tail. The dog’s possible intentions (for example, receiving a pat on the head), feelings, and awareness of self become the focus of the investigation.

With respect to ontogeny , the question becomes: How does the dog’s tail-wagging behaviour develop? The focus here is on investigating the underlying developmental mechanisms that lead to the occurrence of the behaviour. The answer derives from understanding how the sensory-motor mechanisms producing the behaviour are shaped as the dog matures from a puppy into a functional adult animal. Both internal and external factors can shape the behavioral machinery, so understanding the development of the dog’s tail-wagging behaviour requires investigating the influence of the dog’s genes and its experiences.

With respect to function : How does the dog’s tail-wagging behaviour contribute to genetic success? The focus of this question is rooted in the subfield called behavioral ecology; the answer requires investigating the effects of tail wagging on the dog’s survival and reproduction (that is, determining how the tail-wagging behaviour helps the dog survive to adulthood, mate, and rear young in order to perpetuate its genes).

Lastly, with respect to evolutionary history , the question becomes: How did tail-wagging behaviour evolve from its ancestral form to its present form? To address this question, scientists must hypothesize evolutionary antecedent behaviours in ancestral species and attempt to reconstruct the sequence of events over evolutionary time that led from the origin of the trait to the one observed today. For example, an antecedent behaviour to tail wagging by dogs might be tail-raising and tail-vibrating behaviours in ancestral wolves. Perhaps when a prey animal was sighted, such behaviours were used to signal other pack members that a chase was about to begin.

Both the biological and the physical sciences seek explanations of natural phenomena in physicochemical terms. The biological sciences (which include the study of behaviour), however, have an extra dimension relative to the physical sciences. In biology, physicochemical explanations are addressed by Tinbergen’s questions on causation and ontogeny, which taken together are known as “proximate” causes. The extra dimension of biology seeks explanations of biological phenomena in terms of function and evolutionary history, which together are known as “ultimate” causes. In biology, it is legitimate to ask questions concerning the use of this life process today (its function) and how it came to be over geologic time (its evolutionary history). More specifically, the words use and came to be are applied in special ways, namely “promoting genetic success” and “evolved by means of natural selection.” In physics and chemistry, these types of questions are out of bounds. For example, questions concerning the use of the movements of a dog’s tail are reasonable, whereas questions regarding the use of the movements of an ocean’s tides are more metaphysical .

Animal Cognition

The philosophical issues that relate to research on animal cognition can be categorized into three groups: questions about the assumptions on which the research is based; questions that arise about the methods used in the research programs; and questions that arise from the specific research programs themselves.

While the study of animal cognition is largely an empirical endeavor, the practice of science in this area relies on theoretical arguments and assumptions regarding the nature of mind and rationality. If nonhuman animals don't have beliefs, and if all cognitive systems have beliefs, then animals wouldn't be the proper subject of cognitive studies. If animals aren't agents because their behavior isn't caused by propositional attitudes, and if all cognitive systems are agents, we get the same conclusion. While there are arguments against animal minds, the cognitive scientists studying animals largely accept that animals are minded, cognitive systems. Animal consciousness , however, is a different matter; like the topic of human consciousness, it is an area that some scientists are less willing to engage with.

Many of the research programs investigating particular cognitive capacities in different species raise philosophical questions and have implications for philosophical theories, insofar as they impose additional empirical constraints for naturalistically minded philosophers. Traditional research paradigms in animal cognition are similar to those in human cognition, and include an examination of perception, learning, categorization, memory, spatial cognition, numerocity, communication, language, social cognition, theory of mind, causal reasoning, and metacognition.

These different research programs are not always investigated using the same methods. For example, social cognition could be studied by a developmentalist who documents mutual gaze in mother-infant dyads across primate species (e.g. Matsuzawa 2006b), an ethologist interested in free-ranging canid social play behavior (e.g. Bekoff 2001), an experimental psychologist testing theory of mind in an adult symbol-trained chimpanzee (e.g. Premack & Woodruff 1978), an anthropologist observing social games in capuchin monkey communities (e.g. Perry et al. 2003), or a cognitive neuroscientist investigating neural basis of gaze-following in primates (e.g. Emery 2000). Different methods for investigating a phenomenon are based in part on whether the study is in captivity or in the field, and whether developmental stage, social relations, ecology, and past history are taken into account.

1. What is Animal Cognition?

2.1 animal minds, 2.2 rationality, beliefs, and concepts, 2.3 anthropomorphism, 3.1 anecdotal method, 3.2 experimental method, 3.3 ethology, 3.4 comparative cognitive neuroscience, 4.1 communication, 4.2 theory of mind and metacognition, 4.3 emotion and empathy, 5. animal cognition and philosophy: what next, bibliography.

• Other Resources

Related Entries

Animal cognition is the study of the processes used to generate adaptive or flexible behavior in different animal species. As a part of cognitive science , research in animal cognition aims to uncover the different cognitive mechanisms at play across species in order to understand the varieties of cognition, the similarities between humans and other species, and the evolution of cognitive processes.

As with human cognition, there are competing theories about the structure of animal cognition at play. After the cognitive revolution many animal cognition researchers believed that cognition is a general purpose computational system. This was especially true of those who advocated the Social (or Machiavellian) Intelligence Hypothesis (Humphrey 1978; Jolly 1966; Byrne & Whiten 1988; Dunbar 1998). According to the hypothesis, the relatively sophisticated general problem-solving capacities of social animals are due to challenges arising from social living. This view has been challenged by those who claim that the research on other species suggests that cognition is more modular than the traditional view allows. Some argue that individuals cannot use one mechanism to solve all cognitive tasks, since there are different rates of learning and different computational processes implicated in different domains. Because of these differences, cognition must be thought of as a number of special-purpose computational modules rather than one general purpose processor (Hauser & Carey 1998; Shettleworth 1998; Cosmides & Tooby 1994). Other researchers examine animal cognition from within the framework of embodied cognition (Barrett & Henzi 2005), or dynamical systems theory (King 2004; Shanker & King 2002).

Animal cognition research has historically accepted a close relationship between affect and cognition. Early experimental psychologists manipulated subjects' motivation (e.g. by withholding food) and the pain associated with shocks were assumed to cause an unpleasant affective response. Today there is growing interest in the emotional responses of animals, as will be discussed in Section 4.3. In this sense, animal cognition research has anticipated the recent interest in emotion in cognitive science.

2. Theoretical Issues

Philosophical discussion about animal cognition has traditionally focused on the metaphysics and epistemology of mind in creatures who do not have language. While the traditional debate about the existence of animal minds is problematic given the lack of clarity about the nature of mind, recent discussions of animal beliefs and rationality help to make the discussion less muddy.

The early history of western philosophy reflects a tendency to see animals as lacking rationality. Aristotle defined “human” as “the rational animal”, thus rejecting the possibility that any other species is rational (Aristotle Metaphysics ). Aquinas believed that animals are irrational because they are not free (Aquinas Summa Theologica ). Centuries later Descartes defended a distinction between humans and animals based on the belief that language is a necessary condition for mind; on his view animals are soulless machines (Descartes Discourse on the Method ). Locke agreed that animals cannot think, because words are necessary for comprehending universals (Locke Essay Concerning Human Understanding ). Following in this tradition, Kant concludes that since they cannot think about themselves, animals are not rational agents and hence have only instrumental value (Kant Lectures on Ethics ).

However, there were dissenters. Voltaire criticized Descartes' view that humans but not animals have souls and hence minds, by suggesting there is no evidence for the claim (Voltaire Philosophical Dictionary ). Hume was downright dismissive of the animal mind skeptics when he wrote “Next to the ridicule of denying an evident truth, is that of taking much pains to defend it; and no truth appears to me more evident than that beasts are endowed with thought and reason as well as man. The arguments are in this case so obvious, that they never escape the most stupid and ignorant” (Hume Treatise of Human Nature , 176).

Despite Hume's judgment about their worth, much ink has been spilled developing arguments for the existence of animal minds. The two standard arguments are extensions of the two arguments for other minds : the argument from analogy and the inference to the best explanation argument . The argument from analogy for animal minds can be formulated as:

• All animals I already know to have a mind (i.e. humans) have property x .
• Individuals of species y have property x .
• Therefore, individual of species y probably have a mind.

Versions of the argument differ as to what they choose as the reference property x , and how they defend the choice of reference property. The reference property might refer to some general capacity (e.g. problem solving), some specific ability (e.g. language, theory of mind, tool-use), or it might consist of some set of properties.

The argument from analogy for animal minds is in one sense stronger than the argument for other minds insofar as the reference class is larger; rather than starting with the introspection of one's own mind and then generalizing to all other humans, the argument for animal minds takes as given that all humans have minds and generalizes from the human species to other species. In another sense, the analogical argument for animal minds is weaker, since the strength of the argument is a function of the degree of similarity between the reference class and the target class. Humans might be more similar to one another than they are to members of another species.

The second standard argument for animal minds rests on the claim that the existence of animal minds is a better explanation of animal behavior and physiology than those offered by other hypotheses. This argument can be formulated as:

• Individuals of species x engage in behaviors y .
• The best scientific explanation for an individual engaging in behaviors y is that it has a mind.
• Therefore, it is likely that individuals of species x have minds.

The inference to the best explanation argument justifies the attribution of mental states to animals based on the robust predictive and explanatory power that is gained from such attributions. As the argument goes, without such attributions we would be unable to make sense of animal behavior. This argument makes use of standard methods of scientific reasoning; of two hypotheses, the one that better accounts for the phenomenon is the one to be preferred. Those who offer this sort of argument for animal minds are claiming that behaviorist or other mechanistic explanations for animal behavior fail to account for the diversity and flexibility of behavior in at least some species of animal. Critics of this argument offer alternative explanations for the relevant behaviors.

Though philosophers and psychologists seem to generally accept that other species have minds, there is widespread disagreement about what exactly that means. For some, it merely means that some species can feel pain (and hence there are codes of conduct for working with animals). For others, it means that some species are rational agents who have reasons for action. The theoretical investigation into animal cognition examines whether any other species are rational, have beliefs, or have concepts. There are corresponding epistemological questions about how we might ever know the content of an animal's belief or reason for action, given the difficulty of attributing content to a creature who doesn't use language. Whether animals are rational is related to philosophical questions having to do with the moral status of animals . For example, arguments in favor of personhood for great apes are made on the basis of the rational capacities of these species (Cavalieri & Singer 1994), as are arguments for duties toward animals (Skidmore 2001), and arguments given in favor of moral agency (or proto-moral agency) (de Waal 2006; Hauser 2006).

Discussions of animal rationality are confounded by the lack of consensus on what is required for rationality. Because there are many different kinds of rationality (e.g. practical vs. theoretical, process vs. product), disagreements about what sorts of cognitive mechanisms are implicated in rationality (e.g. linguistic processing, logical reasoning, causal reasoning, simulation, biases and heuristics), and disagreements about the extent to which different kinds of normativity are implicated in rationality (e.g. biological fitness or reason-respecting propositional attitudes), there is no straightforward way to answer the question about whether other species are rational agents. Some philosophical theories of rationality are based on an initial acceptance of rationality across species, given evolutionary considerations. For example, on Fred Dretske's view, even some simple learned behaviors, such as a bird's avoiding eating a monarch butterfly, can be construed as minimally rational. Because monarchs who eat toxic milkweed become toxic to birds and other predators, when a bird learns not to eat monarch butterflies after having formed an association between eating monarchs and vomiting, it has a reason for its avoidance behavior. The birds also have a reason to avoid eating a viceroy, given that it is visually almost indistinguishable from a monarch, though not poisonous. The behavior in both cases is explained by the content of the bird's thought (or “thought”), and for Dretske this is sufficient for the bird to count as a minimally rational agent (1988, 2006). Other theories of rationality that take evolutionary considerations into account include those of Daniel Dennett (1995, 1987), Ruth Millikan (2004, 2006), and Joelle Proust (1999, 2006). Another method used to develop theories of rationality is to base it on the human model, and then attempt to extend it to other species. This approach is exemplified by José Bermúdez (2003, 2006). A third class of theories try to stake a middle ground between these two strategies (e.g. Hurley 2003a, 2003b, 2006). For a collection of essays on rationality across species see Hurley & Nudds (2006).

Given the lack of theoretical consensus on the nature of rationality, empirical research projects are not designed to examine rationality directly. Instead, researchers investigate various capabilities that may be associated with rationality. For example, tool-use has long been considered to be an indicator of rational thought. Because tool use involves finding or constructing an object that is utilized as an extension of the body to achieve a goal, it is thought that tool use requires identifying a problem, considering ways of solving the problem, and realizing that other objects can be used in the manipulation of the situation. Early experimental research on chimpanzee problem solving by the German psychologist Wolfgang Köhler had chimps constructing tools to acquire out of reach objects; it was reported that chimpanzees would stack boxes or put together tubes to form a long rod in order to reach bananas hung overhead (Köhler 1925). Given this behavior, Köhler suggested that chimpanzees solve some problems not by trial-and-error or stimulus response association, but through a flash of insight. (But see Povinelli (2000) for a critique of the contemporary interpretation of Köhler's research). Since the days of Köhler, tool use in the wild has been discovered in a number of different taxa, including all great apes, some monkeys, some birds, sea otters, and cetaceans.

Some theoretical arguments about animal rationality identify rationality with other properties, such as having beliefs or concepts. Donald Davidson has offered an argument against animal rationality based on an association between concepts, beliefs, and language. On Davidson's view, believers must have the concept of belief, because to have a belief they must recognize that beliefs can be true or false, and one cannot understand objective truth without understanding the nature of beliefs. In order to develop an understanding of objective truth, one must be able to triangulate with others, to talk to others about the world, and hence all believers must be language users. Since other species lack language, they do not have beliefs (Davidson 1982). Davidson also argues against animal beliefs based on the claim that having a notion of error is necessary for being a believer (Davidson 1975).

A different argument against animal belief has been presented by Stephen Stich, who argues that we cannot attribute propositional attitudes to animals in any metaphysically robust sense, given our inability to attribute content to an animal's purported belief (Stich 1978). On Stich's view, if attribution of belief to animals is understood purely instrumentally, then animals have beliefs. However, if attribution of beliefs to animals requires that we can accurately describe the content of those beliefs, then animals don't have beliefs. Given the second sense of having belief, Stich argues that because “nothing we could discover would enable us to attribute content to an animal's belief” (Stich 1978, 23), we are unable to make de dicto attributions to other species, and we cannot make de re attributions because this would violate the truth-preserving role of attribution. Hence we can make no attribution, and if we can't say what an animal's belief is about, it makes no sense to say that an animal has a belief. The worry here is similar to the worry about anthropomorphism; when we use our language to ascribe content to other species, we may be attributing to them more than is appropriate. Stich is concerned that when we say “Fido believes there is a meaty bone buried in the backyard” we are attributing to Fido concepts he cannot possibility have, concepts like “backyard” which are only comprehensible if one has corresponding concepts such as “property line”, “house”, “fence”, and so on. Stich's argument can be formulated as:

• In order for something to have a belief, it must have a concept.
• In order to have a concept, one must have particular kinds of knowledge, including knowledge of how the concept relates to other concepts.
• Non-human animals don't have this knowledge.
• Therefore, non-human animals don't have beliefs.

While animal cognition researchers agree that we ought to be careful about the concepts that we attribute to other species, many deny Stich's claim that empirical research cannot help us learn anything about the conceptual organization of other species. One of the earliest attempts to examine animal concepts came out of a series of experiments with pigeons. The subjects were shown photographs, and were trained to peck at the pictures that contained a target object, such as a tree, and not respond to the pictures that didn't contain the target object. After the training period, the pigeons were able to generalize to new photographs, pecking only at those that contained trees just as in the training set. It was suggested that this sorting ability demonstrates that the pigeon has a concept of the target object (Herrnstein 1979; Herrnstein et al. 1976).

Many reject the idea that being able to sort objects is sufficient for having a concept corresponding to the sortals. For one, some think language is necessary for concept acquisition (Chater & Heyes 1994). Others think that while concept acquisition is independent of language, sorting behavior alone doesn't demonstrate having a concept, because humans can be trained to sort objects while lacking the corresponding concept. As Colin Allen and Marc Hauser put it, “It is possible to teach a human being to sort distributors from other parts of car engines based on a family resemblance between shapes of distributors. But this ability would not be enough for us to want to say that the person has the concept of a distributor” (Allen & Hauser 1996, 51).

Rather than identifying concept acquisition with sorting behavior, Allen and Hauser suggest alternative methodologies for identifying concepts in other species. For example, they offer a possible (though, they admit, ethically untenable) test for a death concept among vervet monkeys (Allen & Hauser 1996). Vervet mothers are capable of recognizing the alarm cries of their infants, and when they hear such a cry the mother will look towards her infant, and the other females will look towards the mother. Allen and Hauser suggest that playing a recording of a recently deceased infant's alarm cry would help to determine whether vervets have a concept of death. If the mother responds to these recordings in an atypical fashion, unlike the usual response made to a living infant, that response provides evidence of a death concept. According to Allen and Hauser, having a concept permits different responses to identical stimuli. The actual sound of the infant's alarm cry during life is identical to the sound played back after death. If the response to this stimulus is different, this is evidence that there has been a conceptual change associated with the stimulus. Allen presents the general strategy for attributing concepts to animals as follows: “An organism O may reasonably be attributed a concept of X (e.g. TREE) whenever:

• O systematically discriminates some X s from some non- X s; and
• O is capable of detecting some of its own discrimination errors between X s and non- X s; and
• O is capable of learning to better discriminate X s from non- X s as a consequence of its capacity” (Allen 1999, 37).

One way to study the conceptual structure of other species is to use the same methods that are used to study concepts in another group that lacks language, namely human infants (Hauser et al. 1996; Hauser & Carey 1998; Bermúdez 2003; Gómez 2005). The preferential looking time paradigm, also known as the dishabituation paradigm, is used to study human infants' understanding of the physical and social world (Baillargeon & DeVos 1991; Spelke 1991). Dishabituation experiments are thought to help us understand what kinds of predictions infants make about their word, and this information can help us determine how they see the world. The methodology is simple; an infant is repeatedly shown a stimulus, and after becoming habituated to the stimulus the infant becomes disinterested. At this point, a new stimulus is shown. If the infant sees the new stimulus as different from the target stimulus, or impossible given the target stimulus, the infant will look longer at the new stimulus. If the infant takes the new stimulus to be similar to the target stimulus, then she will not show any additional interest. The idea is that by comparing responses among groups of individuals, a researcher can learn something about how that group conceptualizes the world.

In one study using this method, Marc Hauser and colleagues investigated numerical concepts in different primate species, including rhesus monkeys (Hauser et al. 1996) and cotton-top tamarins (Uller 1997). The researchers tested the monkeys' ability to keep track of individual objects placed behind a barrier. They found that like human infants, the monkeys look longer at impossible outcomes. For example, in one test condition the rhesus monkeys were shown two eggplants serially placed behind a screen, and then the screen was lifted showing only one eggplant. The monkeys looked longer at the one eggplant than they did when shown the expected two eggplants, suggesting that they represent the eggplants as distinct sortals.

Another way we might learn how different species organize the world is to teach individuals a symbolic communication system. For example, the biologist Irene Pepperberg trained an African Grey parrot named Alex to sort objects using meta-level concepts that categorize other concepts. Alex was able to sort objects according to color, shape, and matter, and he was able to sort sets of objects according to number. In addition to sorting, Alex could report which feature makes two objects similar or different. For example, when presented with a red block and a red key, Alex responded to the question “What's same?” by uttering “color.” He could also report similarities and differences in shape and matter. Pepperberg claims that her studies demonstrate Alex's understanding of categorical concepts, and reveal the classifications that Alex devised (Pepperberg 1999). However, one might be worried that the concepts exhibited by symbol-trained animals are an artifact of the communication system, and not typical of the species.

When researchers attribute mental states to other species, they open themselves to the charge of anthropomorphism. The term “anthropomorphism” refers to the act of attributing uniquely human traits to other animals; the traits in question are usually psychological states. In recent years, there have been a number of theoretical discussions about the charge of anthropomorphism itself (including the essays in Mitchell et al. 1997; Daston & Mitman 2005; and work by Fisher 1990, 1991; Kennedy 1992; Crist 1999; Rivas & Burghardt 2002; Keeley 2004; Andrews forthcoming).

In response to charges that psychological and agential attributions are examples of anthropomorphic attribution, some have argued that the charge of anthropomorphism is a charge that the attributor is making a category mistake, rather than merely a false attribution. It is a claim that the attribution must be logically false, because members of the target species are not the sorts of things to which the term can apply (Fisher 1991; Keeley 2004). However, if the charge of anthropomorphism is the charge that the attributer is making a category mistake, then the charge is being made on conceptual, rather than empirical grounds; hence the worry that refusing to attribute so-called anthropomorphic properties without first examining whether they might be held by members of any other species is unscientific (Fisher 1990; Asquith 1997; Keeley 2004). Thus, one response to the charge of anthropomorphism is continued research, for one wouldn't know whether a property is anthropomorphic until after the relevant research has been completed. As Sober puts it, “The only prophylactic we need is empiricism” (Sober 2005, 97).

However, Sober also argues that the empirical methodology of psychology places a different burden of proof on animal cognition and human cognition research. This is because the null hypothesis in the animal cognition research is that there are different cognitive mechanisms at work in humans and animals. Given that type 1 errors (reporting a false positive and rejecting a (possibly true) null hypothesis) are taken to be more serious errors than are type 2 errors (reporting a false negative and not rejecting a null hypothesis when it should be rejected), the practice of science results in a bias against attributing psychological traits to animals (Sober 2005). The debate about how to interpret the results of animal studies as compared to human studies may be seen as a debate about an inconsistent application of Morgan's Canon. Morgan's Canon states: “In no case may we interpret an action as the outcome of the exercise of a higher psychical faculty, if it can be interpreted as the outcome of the exercise of one which stands lower in the psychological scale” (Morgan 1894, 53). Though this is a longstanding rule of thumb in animal cognition research, sometimes referred to as the “principle of conservatism,” it is not a principle commonly used in human cognition research. To complicate matters, attempts to determine what exactly Morgan's Canon instructs a researcher to do have raised worries about its meaningfulness (Sober 2005; Allen-Hermanson 2005).

Despite the defenses given for attributing mental states to animals, worries about anthropomorphism remain. Kennedy claims that the arguments for attributing mental properties to animals often rest on a false dichotomy: either animals are stimulus-response machines, or they are agents with beliefs and desires. Since animals are not stimulus-response machines, they must be psychological agents (Kennedy 1992). The problem with this argument is that not all machines implement stimulus-response functions; some machines are complex and indeterministic, and if animals were machines, they would be machines of that sort (Barlow 1990; Kennedy 1992).

Other critics rely on arguments much like those of Stich and Davidson discussed above. If we have no good scientific methods for attributing mental states to creatures without language, then we should not make such attributions. Since we are barred from making the attributions, scientific psychology ought not engage in analyzing animal mentality (e.g. Keeton 1967; Kennedy 1992; Blumberg & Wasserman 1995). Anthropomorphism is seen as a human tendency that must be overcome in order to do good science.

It has been noted that such arguments are about the proper methods of science, the scope of science, and how to interpret data (Keeley 2004; Bekoff & Allen 1997). As such, the argument is not an empirical one, but a theoretical one. This can be seen in the way the debates sometimes result in an impasse. Those opposed to attributing mental properties to animals are accused of begging the question (Griffin 1992), by committing “reverse anthropocentrism” (Sheets-Johnstone 1992) or “anthropodenial” (de Waal 1999). The charge of begging the question goes both ways. Kennedy argues that arguments for animal mentality are grounded in human intuition or introspection and that introspection is itself anthropomorphic and ought not be taken as objective evidence (Kennedy 1992). As both sides accuse the other of begging the question, some conclude that the debate is not fecund, and ought to be replaced with empirical work in comparative biology and psychology (Keeley 2004; Sober 2005; Andrews forthcoming).

The concerns about anthropomorphism appear to be largely limited to western scientists. It has been argued that researchers from countries with a Buddhist rather than Christian orientation are not culturally encouraged to see a categorical distinction between humans and nonhuman animals (Asquith 1991; Sakura 1998; Matsuzawa 2003; de Waal 2003). Unlike Christianity, Buddhist doctrine does not claim that humans but not animals have immortal souls, and it does not allow humans to use animals for their own purposes in the ways Christianity does. The Buddhist tradition sees a connection between humans and other animals, and allows that humans can be reborn as animals. De Waal argues that the difference in cultural attitudes toward animals led to an early rejection of Japanese methods and findings in primatology, and that it is only recently that some of those ideas, such as Kinji Imanishi's claim that primates display cultural differences within species, have made their way into western scientific discourse (de Waal 2003).

3. The Science of Animal Cognition

Scientific interest in animal minds and cognitive capacities grew as a result of Charles Darwin's theory of evolution by natural selection. In The Descent of Man , Darwin introduced many of the issues that motivate the research programs in animal cognition today, including tool use, reasoning, learning, concepts, consciousness, the social sense, and the moral sense. He was also interested in animals' aesthetic judgments and whether they believe in the supernatural, issues that haven't been taken up by contemporary researchers. In addition, Darwin anticipated current interest in implicit reasoning with his comment “The savage would certainly neither know nor care by what law the desired movements were effected; yet his act would be guided by a rude process of reasoning, as surely as would a philosopher in his longest chain of deductions” (Darwin 1974, 75)

Darwin advocated the continuity of the mental across species; just as some morphological characteristics are homologous across species living in similar environments, we should expect psychological and behavioral similarities as well: “the difference in mind between man and the higher animals, great as it is, certainly is one of degree and not of kind” (Darwin 1974, 122). This view was also advocated by Darwin's contemporary, the naturalist George Romanes, who in his book Animal Intelligence writes “there must be a psychological, no less than a physiological, continuity extending throughout the length and breadth of the animal kingdom” (Romanes 1970, 10).

The method that Darwin, Romanes, and other contemporaries used to investigate these questions could be described as the anecdotal method. Stories about animal behavior were collected from a variety of people, including military officers, amateur naturalists, and layfolk, and were compiled and used as evidence for a particular cognitive capacity in that species.

The anecdotal method as practiced by Darwin and Romanes has been criticized for a number of reasons. The “evidence” gathered was often a story told about an event witnessed by a single person, usually not a trained scientific observer. In addition, these stories were often acquired second- or third- hand, so there were worries that the reports had been embellished or otherwise altered along the way. These problems were recognized early on, and in response Romanes developed three principles for accepting anecdotes in order to avoid some of these problems:

• Never accept an incident report as fact without considering the authority or respectability of the observer.
• If the observer isn't known, and the incident report is sufficiently important, consider whether the observer may have reason or cause to make an inaccurate report.
• Look for corroborations of the observation by examining similar or analogous observations made by other independent observers (Romanes 1970).

The third principle was the one he most relied on, writing “This principle I have found to be a great use in guiding my selection of instances, for where statements of fact which present nothing intrinsically improbable are found to be unconsciously confirmed by different observers, they have as good a right to be deemed trustworthy as statements which stand on the single authority of a known observer, and I have found the former to be at least as abundant as the latter” (Romanes 1970, ix).

Despite Romanes' attempts, the method remained problematic insofar as it didn't provide any statistical information about the frequency of such behaviors; selection bias would lead people to report only the interesting intelligent behaviors and ignore the frequency of behaviors that might serve as counterevidence. Thus, the anecdotal method as practiced by Darwin and Romanes fails many of the virtues associated with good scientific method.

The legacy of Darwin and Romanes for animal cognition can be summed up in the oft-quoted phrase “The plural of ‘anecdote’ is not ‘data’.” Though 19 th century anecdotalism has been rejected, defenses of the use of observations (often now referred to as “incidents” or “qualitative reports” rather than “anecdotes”) come from ethology. Among ethologists, collecting behavioral incidents is part of the standard methodology, and fieldwork by trained scientists has resulted in greater knowledge of animal behavior. For example, when Jane Goodall reported seeing chimpanzee hunting and lethal intergroup aggression, the scientific image of chimpanzees had to be significantly revised (Goodall 1986).

A middle ground in the use of qualitative reports takes the collection of incidents as a research tool. To be useful as a research tool, the concerns about anecdotalism must be addressed. One major concern has to do with the interpretation and description of the behavior as it is reported. Like concerns about attributing content to animal minds, the worry is that there will be an over-attribution due to the bias of the scientist. In addition, if anecdotes are presented as stories, and the story structure is used to determine truth, there is a worry that such stories will be elaborated so as to remove inconsistencies or fill gaps (Mitchell 1997). The psychologist Richard Byrne defends the scientific use of rare events as a useful tool, writing, “careful and unbiased recording of unanticipated or rare events, followed by collation and an attempt at systematic analysis, cannot be harmful. At worst, the exercise will be superseded and made redundant by methods that give greater control; at best, the collated data may become important to theory” (Byrne 1997, 135). It is important to note that Byrne is not talking just about data acquired by the ethological method, but also incidents observed by scientist in the field or lab who are well-versed in the baseline behavior of the species.

Before Watson and Skinner promoted behaviorism in human psychology, similar ideas were being espoused among animal cognition researchers who were disappointed in the lack of rigorous methods for studying animal minds. The psychologist C. Lloyd Morgan's early text on comparative cognition was critical of the anthropomorphism in the anecdotes reported by Darwin, Romanes, and others. In An Introduction to Comparative Psychology (1894), Morgan famously proposed what is now called Morgan's Canon; recall that Morgan's Canon is a principle of conservatism that instructs researchers to interpret behavior as caused by the lowest possible “psychical faculty” (Morgan 1894, 53). In his text Morgan suggested that many of the seemingly cognitively sophisticated behaviors of animals could be explained by associative learning.

The Clever Hans scandal of 1904 demonstrated Morgan's Canon in use; Hans was a famous Russian trotting horse who charmed crowds by appearing to calculate mathematical problems, as well as to read German and musical notation, simply by tapping his hoof (Candland 1993; Pfungst 1965). After much investigation, the experimental psychologist Otto Pfungst found that Hans wasn't counting or reading language; rather he was reading his owner von Osten's bodily motions. Von Osten was unconsciously cuing Hans to start and stop tapping his foot at the correct time, and Hans had merely leaned to associate von Osten's movements with the correct behavior. Today, the legacy of Clever Hans can be seen in the controls used during experimental tests of an animal's ability. For example, researchers who know the correct response will wear a welder's mask, blackened goggles, or some other device to keep the subject from being cued by eye gaze or facial expressions, or naive trainers are used during testing.

One of the leading animal cognition researchers during this time was Edward L. Thorndike (1874-1949). Thorndike argued for the necessity of experimental study of animal intelligence, writing, “most of the books do not give us a psychology, but rather a eulogy of animals. They have all been about animal intelligence , never about animal stupidity” (Thorndike 1911, 22). Experiments will help us learn both what animals can do and what they fail to do, thus giving us a better overall understanding of animal cognition. One of Thorndike's projects focused on the problem solving abilities of housecats. He placed cats in a variety of puzzle boxes, and observed the strategies the cats used to escape. When first put in a new box, the cats took a long time to find the solution, but after experience with the box they were able to escape much more quickly. Thorndike found that the cats improved by ignoring the ineffective actions and performing the useful ones. This suggested to Thorndike that cats learn through trial and error, and his conclusion helped to reinforce the belief that animal behavior can be fully explained in associative terms.

While the behaviorists succeeded in introducing much-needed rigor into the study of animal minds, there was some concern that they had gone too far, that the methods were too stringent, and that the drive for repeatable and controlled experiments could not be used to uncover all there is to know about the function of animal minds. For example, the ethologists thought that to understand animal behavior, animals must be observed in their natural environment. As sterile laboratory experiments stripped of social and environmental context, some considered the behaviorists' studies ecologically invalid.

While today some experimentalists defend these methods (Povinelli & Vonk 2004), other experimentalists agree with the criticisms, and in response have developed methods that are sensitive not only to environmental concerns, but also to development and social relations. For example, research coming out of Kyoto University's Primate Research Institute (PRI) is based on a three-part research program (Matsuzawa et al. 2006). First, chimpanzee physical, cognitive, and social development is taken into account in the design of experiments, and subjects are raised by their mothers rather than by human caregivers or unrelated animals. In addition, lab work and fieldwork is synthesized; field observations are used to develop experiments, and experiments are conducted both in the field and in the laboratory. Finally, the method includes analysis of the physiological and biological features of the species that could be related to cognitive abilities.

The experimental research at PRI uses what they call the “participant observation” method, which is based on the triadic social relationship between mother, infant, and experimenter. When testing chimpanzees in the lab, they are never taken from their natural social environment; rather, the experiments are brought into the social environment. As a researcher becomes a member of that social environment, she can run experiments that are woven into normal daily activities. At PRI, a different researcher is bonded with each mother-infant dyad, and the relationship is expected to last a lifetime. This close relationship between human and chimpanzee is thought to offer many benefits. It makes the chimpanzees more willing to engage in the research activities, so researcher can better understand of what chimpanzees can and cannot do (rather than what they will and won't do). In addition, Matsuzawa claims that the participant observation method is better at investigating species typical social cognition than are isolated experiments on single subjects, because the PRI chimpanzees are not integrated into a human social environment, but the researchers adapt to the chimpanzee social environment (Matsuzawa 2006a). Finally, the bond between researcher and subject allows the human to interact with his chimpanzee “research partners” at a younger age, given the trust between researcher and mother. Mother and infant can be taught a task together, which can help to illuminate developmental differences in particular abilities. For example, Inoue and Matsuzawa have recently reported that infant chimpanzees are better able to recall strings of numerals in order than are adult chimpanzees and humans (Inoue & Matsuzawa 2007).

While the behaviorists were performing laboratory experiments, ethologists such as Oskar Heinroth, Konrad Lorenz, Nikolaas Tinbergen, and Karl von Frisch were following animals into the field to observe naturally occurring behavior. The ethological method is based in biology, and takes into account not only the behavior, but also the context of behavior, the environment, and the physiology and evolutionary history of the animal.

Lorenz and Tinbergen were interested in analyzing the complex and rigid set of movements that make up a single act. They postulated that such movements, which they called fixed action patterns, are innate and caused by the existence of a releasing mechanism that responds to some external sensory stimulus. Ethologists study such acts at various organizational levels, e.g. at the individual, dyad, family group, and species level (Menzel 1969). To explain behavior, ethologists follow Tinbergen's suggestion that we can distinguish between explanations in terms of proximate causes , such as mechanism or function, and ultimate causes , such as ontogeny (the maturational processes involved in the behavior) and evolution (Tinbergen 1963).

But before an explanation for some behavior can be found, the behavior must be well understood in the context of species normal behavior. Thus, the ethologist will begin the study of a species by constructing an ethogram from field notes taken after many hours of observation (Brown 1975; Lehner 1996). An ethogram is a thorough catalogue of the characteristic behavioral units of a species, and each unit is given a verbal description and perhaps an image. Theoretical debates arise over the issue of how an ethogram should label and describe behavioral units. Behaviors can be described formally, and describe the action at the level of muscular contractions, or patterns of bodily movements (e.g. beak pecks ground). On the other hand, an ethogram could describe behaviors functionally, and place the behavior in a larger context by referring to the purpose or consequence of the behavior (e.g. eat) (Hinde 1970).

There are criticisms of both functional and formal methods of description. Formal descriptions may leave out important aspects of an animal's behavior, whereas functional descriptions are subject to over-interpretation and may lead to anthropomorphism, and they may conflate explanations in terms of ultimate and proximate reasons (see Allen & Bekoff 1997 for a discussion). Millikan (1993) and Allen and Bekoff (1997) provide philosophical defenses of relying on functional descriptions in ethology. While Millikan has claimed that ethologists should only be concerned with behavior as functionally described, Allen and Bekoff argue that the choice between a functional and formal description will vary on the context, depending on which is more useful. In many cases, functional descriptions will be preferred because of the advantages Hinde (1970) has identified. For one, behavior described functionally will result in fewer data sets, making for more robust data analysis. In addition, descriptions in terms of function are more informative than formal ones, given that they include information about the cause of the behavior or its consequence. Finally, behavioral changes can be described in terms of environmental changes.

However an ethologist decides to describe behaviors, the question of how to individuate a behavior arises (Skinner 1935; Russon et al. 2007). Descriptions of behavior can be finely grained, and refer to the specifics of a behavior, e.g. using a stone (or a leaf cup, or chewed leaves, or a hand, or fur, etc.) for drinking water from a river. On the other hand, behaviors can be roughly grained into larger behavioral units, e.g. using a drinking tool. If behaviors should be categorized to reflect the way the species organizes its behavior, then identifying behaviors requires first knowing what the species' internal organizational scheme is (Byrne 1999; Russon et al. 2007).

In addition to classical ethology, there is a field of study called cognitive ethology, founded by psychologist Donald Griffin. Cognitive ethology purports to study animal consciousness (Griffin 1984, 1985, 1992), which is often identified with the study of beliefs, intentions, self-awareness, deception, and theory of mind (Shettleworth 1998). Griffin's use of the word “consciousness” belies his greater interest in cognition, given that most of the topics he discusses (e.g. perception, memory, spatial cognition, language, tool use etc.) are cognitive (Griffin 1992). While “cognitive ethology” is often used to describe animal cognition research that uses the methodologies of classical ethology, it appears that few researchers welcome the label (Allen 2004).

A more recent methodology for studying animal cognition results from recent technological advances in brain imaging, and many current discussions of cognitive evolution relies on data from comparative neuroscience.

As neural correlates for cognitive abilities have been found in humans, neuroscientists have looked for parallel structures in other species. For example, while language is considered uniquely human, a distinct cortical area in the macaque monkey is thought to be a precursor to Broca's area in humans. Broca's area is necessary for proper speech production in humans, and the analogous area in macaques is used for controlling facial muscles. It is hypothesized that this area has the function of regulating control over facial expressions related to communication (Petrides et al. 2005).

Others argue that the mirror neuron system found in monkeys and humans may be partially homologous to Broca's area (Rizzolatti & Craighero 2004). The mirror neuron system becomes active both when the subject engages in a behavior and when the subject observes another engage in a similar action, and has been suggested as a neuronal basis for abilities such as theory of mind, empathy, and imitation (Goldman 2006).

Comparative cognitive neuroscience plays a significant role in discussions of the Social Intelligence Hypothesis. If large executive brains evolved in response to social challenges, then larger executive brains ought to be positively correlated with increased intelligence (Dunbar 1998). One problem with testing this thesis has been determining what counts as the executive brain. Comparative studies of brain size, relative brain size, encephalization quotient, and brain capacity have met with counterexamples (for example, the shrew's brain is much larger than the human brain, relatively speaking). It has been argued that the number of cortical neurons, and the thickness and density in the cortex is positively correlated to a greater information processing capacity (Roth and Dicke 2005).

4. Research Programs

The research programs in animal cognition are too numerous to thoroughly cover here; good introductory texts can be found that introduce a variety of topics (e.g. Shettleworth 1998; Roberts 1997; Pearce 1997). What follows is a brief introduction to three areas of research that have been of interest to philosophers: communication, theory of mind, and empathy. While these areas are those that have been discussed by philosophers, the choice of these paradigms is not meant to undermine the philosophical importance of other areas of research, such as concept acquisition, perception, causal reasoning, memory, culture, imitation, innovation, and so forth. And a final caveat: these three areas are characterized primarily by research on great apes (chimpanzees, bonobos, gorillas, orangutans, as well as humans), but this emphasis does not imply that philosophical issues don't arise in research with other taxa, or that these cognitive capacities only occur in great apes.

To communicate, an individual must be capable of engaging in some behavior that transmits information to others, and the individual must have some control over whether the behavior is exhibited. Communication is distinguished from signalling; while signaling, like communication, can cause a change in another's behavior, signaling only accomplishes this in an inflexible and rote way. Research on communication focuses both on the natural communication systems of animals in the wild, and on attempts to introduce novel communication systems to animals in a laboratory setting.

4.1.1 Artificial Symbolic Communication Studies

In the 20 th century there was great interest in teaching symbolic communication systems to other species. The earliest forays into this area were with chimpanzees, and focused on teaching spoken language to chimpanzees raised as human children (Kellogg & Kellogg 1933; Hayes & Hayes 1951). With the realization that chimpanzees lack the vocal apparatus needed to form human words, research shifted to teaching chimpanzees American Sign Language and artificial symbolic communication systems. The first such study, Beatrix and Allen Gardner's Project Washoe, was initially reported to be a success. Using explicit training methods, including shaping, molding, and modeling, the researchers were able to train the infant Washoe to form at least 132 ASL signs. Focus was on production of gestures, rather than comprehension, and the Gardners' stated intention was to train Washoe (and later, other chimpanzees) in a social setting, mimicking the language-learning environment of children as much as possible. The Gardners claim that, “[Washoe] learned a natural human language and her early utterances were highly similar to, perhaps indistinguishable from, the early utterances of human children. Now, the categorical question, can a nonhuman being use a human language, must be replaced by quantitative questions: how much human language, how soon, or how far can they go?” (Gardner & Gardner 1978, 73).

While the Gardners' claims about ape language were being echoed by others working on ape language (e.g., Premack 1971; Patterson 1978), not everyone agreed. The psychologist Herbert Terrace, who used the methods of the Gardners to teach ASL signs to an infant chimpanzee named Nim Chimpsky, argued that the apes were not using the signs to communicate. Terrace concluded that some of the results achieved by the Gardners could be explained by associative learning rather than comprehension of the semantics of the symbols. In his study, he tried to control for associative learning, and his focus on syntax had him attend to symbol order in multi-symbol strings. While early results of this study seemed promising, after watching videos of Nim's symbol use he noticed that what had been initially seen as spontaneous utterances were often imitations of utterances just made by his trainers. Terrace reviewed films of Washoe's utterances, and found similar patterns: the teacher initiates the signing, and the chimpanzee mimics the teacher's signs. He also noted that the give and take rhythm of child-adult communication was not mirrored by the chimpanzee-trainer conversations, and took this difference in pragmatics as further evidence that the chimpanzees were not using language (Terrace et al. 1979).

Though the Gardners defended their studies against Terrace's critiques (Gardner & Gardner 1989), other researchers tried to control for alternative interpretations of their results. Premack, for example, relied on transfer tests as evidence that Sara understands the symbols she was taught (Premack 1971). In a transfer test, a new symbol is taught only in the context of a subset of the subject's vocabulary. Once the subject reaches criterion on the teaching set, a formal test is conducted using novel strings of symbols.

The post-Terrace research on symbolic communication has expanded to include different species, such as the other apes, dolphins, parrots, and sea lions. In addition, the focus of some studies has shifted from syntax to semantics, and from production to comprehension.

Advocates of this research program argue that the studies uncover something about the relationship between language and mind, the evolution of human language, and the roles played by development and scaffolding in human language (Lloyd 2004). However, to the critics, these studies are just more evidence of the power of association, and the ability of humans to train animals to do anything. There is a huge literature on these studies, with critics (Pinker & Bloom 1990; Pinker 1994; Chomsky 1980) as well as defenders (Lloyd 2004; Greenfield 1991; Savage Rumbaugh et al. 1998).

One area of contention has to do with whether animals who successfully use some aspect of human language are using it qua language, or are instead engaged in symbolic communication. At least three different demarcations between language and other symbolic communication systems have been offered. According to Noam Chomsky's original linguistic program, to use language is to embody certain structural principles, and all language users are able to produce a potentially infinite number of grammatical strings via recursive embedding (Chomsky 1968). The linguistic anthropologist Charles Hockett identified up to seventeen design features that occur in every human language, including semanticity, discreteness, and arbitrariness (Hockett 1977). More recently Hauser, Fitch, and Chomsky (2002) argue that the mechanism that allows for recursive thinking is the central cognitive requirement for language, and is a feature of human communication systems not found in other species.

Chomsky was a vocal critic of early animal language studies, especially of the claims made by some researchers that the apes had acquired language. For Chomsky, language requires syntax, something that is lacking in all the communication systems of the apes. Furthermore, to train an ape to use symbols is a laborious process, whereas children learn language effortlessly. Language is innate, according to Chomsky, so if apes had the capacity for learning language, they would speak without human intervention (Chomsky 1968). Chomsky often states his criticism as an a priori argument against animal language: “if an animal had a capacity as biologically sophisticated as language but somehow hadn't used it until now, it would be an evolutionary miracle, like finding an island of humans who could be taught to fly” (cited in Lloyd 2004, 585).

Another argument Chomsky has offered against animal language is based on the dissimilarity between animal communication systems and human language. He writes, “The question of whether other systems are ‘like’ human language is a question about the usefulness of a certain metaphor” (Chomsky 1980, 434), and he argues that the structural principles, manner of use, and ontogenetic development of ape symbol use is so different from human language that any analogy between the two would be very weak. Those who defend the animal symbolic communication system as language take Chomsky to task on this point, and stress the similarities between the two systems of communication.

Given recent findings in genetics, the biological capacity for language may be more accurately described as a collection of biological capacities, some of which we share with other species. The FOXP2 gene is found to play a role in speech production, and some claim that it was instrumental in the development of language in humans. The FOXP2 gene is also expressed in the same part of the brain in zebra finches, and it has been reported that finch fledglings with reduced FOXP2 are impaired in their ability to learn to sing (Haesler et al. 2007).

There appears to be less concern about describing animal symbol use as communicative, though many theories also portray communication as carrying large cognitive demands. For example, according to H.P. Grice's classical pragmatics view of communication, for an individual to make a meaningful utterance, the speaker must intend to change the belief of the audience, and the audience must recognize the speaker's intention (Grice 1969). This strong requirement for communication requires metacognition on the part of communicators, including a theory of mind, or the ability to attribute mental states such as beliefs and desires to another. As we will see in section 4.2.1, claims about theory of mind in nonhuman animals are cause for much skepticism. A weaker reading of Grice's criteria might limit the requirement of communication to a recognition of intentionality, something which is rather less controversial.

4.1.2 Natural Communication

In The Descent of Man , Darwin claims that communication can be seen in many species: in the alarm cries of monkeys, the distinctly meaningful barks of dogs, and the songs of birds that that are learned from adult experts (Darwin 1871). Darwin's evidence took the form of anecdotes, but soon formal studies of natural animal communication systems were attempted. One of the earliest projects was that of Richard Lynch Garner, who, convinced that primate vocalizations are linguistic in nature, intended to learn the natural language of the apes. Garner traveled to central Africa in 1892 and lived in a cage in the forest, to protect himself from the chimpanzees and gorillas he was observing. Since the time of Garner's failed research, ethologists, biologists, and psychologists have documented the natural communication systems of hundreds of species.

Though these systems are described as communication, the central theoretical questions are whether the communicative utterances are referential and whether the utterers are mentally representing the referent, that is, whether the utterance is meaningful from the perspective of conspecifics. These questions are applied just as well to the symbolic communication research discussed above.

One possibility is that vocalizations such as alarm calls could be the result of an involuntary reaction to stimuli, an emotional display, or an indication of the severity of the threat. On the other hand, alarm calls might be meaningful signals can be distinguished from involuntary behavioral responses. Those who argue that animal communication systems and human language are homologous (functionally similar due to common evolutionary origin) or analogous (functionally similar with different evolutionary origins) to one another have attempted to demonstrate that some animal signals are referential. One test for referential communication is to see if the behavior is flexible by determining whether there is only a probabilistic relationship between the stimulus and response. For example, if there are different responses to different utterers (e.g. infant vs. adult, dominant vs. submissive), this is thought to demonstrate flexibility in the behavior that is suggestive of referential understanding (Evans 2002; Tomasello & Zuberbühler 2002). In addition, it is thought that referential calls will encode specific information about the predator and that animals who hear the alarm call perceive that encoded information (Evans et al. 1993a, Evans 1997). Marler et al. (1992) offer two criteria that must be met for a signal to be functionally referential. The production criterion requires that all the stimuli that elicit the signal can be said to belong to one category, either a general category such as “aerial predators” or more specific one such as “eagle.” The perception criterion states that utterance of the referential signal is alone sufficient to elicit the same behavior as would be elicited by perceiving the referent (Marler et al. 1992).

Given these criteria, Marler and Evans examined the anti-predator behavior of bantam chickens, and found that the chickens reliably give different alarm calls to aerial predators and ground predators (Evans et al 1993b; Evans & Marler 1995). Because they also behave differently toward the two different predators, Marler and Evans suggest that the alarm cries functionally refer to the kind of predator approaching. When a chicken emits a scream after seeing a hawk, they claim the chicken is referring to the hawk, rather than expressing fear of the hawk, or ordering conspecifics to take cover, crouch, and look up to the sky.

Alarm calls and other communicative vocalizations that fulfill these requirements are found in many species. Gunnison's prairie dogs, for example, give different alarm calls to humans, hawks, and dogs/coyotes. In response to the hawk alarm call, only the prairie dogs that are in the flight path of the hawk respond, running into a burrow. The human alarm call elicits a community wide flight into the burrows, whereas the dog/coyote alarm call leads all individuals to run to the edge of the burrow and stand erect (Kiriazis & Slobodchikoff 2006). Vervet monkeys also give alarm calls for different predators. Following up on the field observations of zoologist Thomas Struhsaker (1967), Dorthy Cheney and Robert Seyfarth used playbacks of prerecorded alarm calls to demonstrate that when a leopard alarm is sounded, the vervets run into trees, where they are safe from the leopards due to the monkeys' agility in jumping from tree to tree. When an eagle alarm is sounded, monkeys look up and run into bushes. When the snake alarm is sounded, the monkeys stand bipedally and peer into the grass around them (Cheney & Seyfarth 1990). Other species found to have different alarm calls and different behavior for different species of predator include Diana monkeys (Zuberbühler 2000), Campbell's monkeys (Zuberbühler 2001), and meerkats (Manser 2001; Manser et al. 2001). In addition, ground squirrels (Owings & Hennessy 1984), tree squirrels (Green & Meagher 1998), and dwarf mongooses (Beynon & Rasa 1989) are all known to have alarm calls that distinguish between terrestrial and aerial predators.

The emphasis on alarm calls isn't to suggest that this is the only area in which other species are said to use referential communication. Many other calls and gestures are thought to involve referential communication. For example, the food calls of chimpanzees are thought to indicate not only the presence of food, but also the location or quality of food (e.g. Slocombe & Zuberbühler 2005, 2006). Similar findings have been reported for chickens (Evans & Evans 2007). Bottlenose dolphins are said to refer to themselves via their signature whistle (Janik et al. 2006). In addition, we shouldn't expect that all communication proceeds through calls or other vocalizations; in some species olfaction or echolocation may be more salient. Allen and Saidel suggest that empirical work can be done to determine the kinds of referential communication different species can engage in, and that such research can help us to better understand the correct description of the mechanisms subsuming human language (Allen & Saidel 1998).

Metacognition is generally understood as cognition about cognitive processes, i.e., thoughts about thoughts. One can engage in metacognition when thinking about one's one thoughts, desires, etc. or by thinking about the mental processes of others. Theory of mind, or mindreading, is associated with the latter ability. Since the late 1970s, philosophers and scientists alike have been pursuing this research program.

4.2.1 Theory of Mind

Like humans, many species are social animals who, in addition to navigating a physical world, must also navigate a social world. In the 1970's it was suggested that in order to succeed in a social world, individuals would benefit from having some understanding of the mind of others. The psychologist Nicholas Humphrey was one of the first to propose that social knowledge requires theoretical knowledge of the causes of behavior (Humphrey 1976, 1978). Humphrey wrote, “...I venture to suggest that if a rat's knowledge of the behaviour of other rats were to be limited to everything which behaviourists have discovered about rats to date, the rat would show so little understanding of its fellows that it would bungle disastrously every social interaction it engaged in; the prospects for a man similarly constrained would be still more dismal” (Humphrey 1978, 60).

The term “theory of mind” was introduced by psychologists David Premack and Guy Woodruff the around the same time, and they make the same assumption regarding social cognition. The specific question Premack and Woodruff were interested in was whether the chimpanzee attributes beliefs and desires in order to predict and explain behavior, something they assumed that humans do. In effect, Premack and Woodruff wanted to know whether a chimpanzee is a Humean action theorist who understands the behavior of others as being caused by propositional attitudes. Thus, they defined theory of mind as “the ability to predict and explain behavior by attributing mental states.” Premack and Woodruff attempted to determine whether Sarah, the same chimpanzee from Premack's symbolic communication project, has a theory of mind. To examine whether Sarah understands what others believe, they used the following paradigm: Sarah was shown videotapes of humans trying to solve certain tasks (e.g. acquiring out of reach bananas, warming up a cold room by lighting a heater) and she was supposed to choose from an array of photos to pick the solution (Premack & Woodruff 1978). Because Sarah picked the correct photograph at an above-chance level, Premack and Woodruff concluded that she has a theory of mind. They claimed that Sarah must have been attributing “at least two states of mind to the human actor, namely, intention or purpose on the one hand, and knowledge or belief on the other” (Premack & Woodruff 1978, 518).

In commentary on this study, it was pointed out that Sarah could have used other methods to solve the problems. She could, for example, have attended to the goal of the actors, as opposed to their mental states (which is the interpretation that Premack now endorses (Premack & Premack 2003)). Most of the commentators were unconvinced by the design of the study, and several suggested alternative methodologies for examining the question. One suggestion was to require the subject to solve a coordination problem. To succeed in a coordination problem the subject would have to alter his own behavior in expectation of what another will do (e.g. Bennett 1978; Dennett 1978). Dennett suggests that a good coordination problem might require that the subject considers another's false belief, so that the behavior that is predicted will be an unusual one, such as a behavior that would only be taken if the actor had a false belief. A false belief coordination problem would avoid alternative interpretations having to do with identifying the actor's goal, or making associations from similar situations in the past. The behavior performed by an actor who has a false belief will not achieve the actor's goal, and will probably not be something the subject has witnessed previously. The main problem with this suggestion, Dennett notes, is how to determine the content of the predictions a chimpanzee might make.

Given the difficulties associated with developing a good nonverbal test for theory of mind (Dennett 1983), Dennett's suggestion was taken up by researchers interested in studying theory of mind in children (Wimmer & Perner 1983). Wimmer and Perner accepted Premack and Woodruff's definition of theory of mind and asked when the small child gains a theory of mind. To answer that question, they designed the false belief task , which was to become a standard test for theory of mind. Children watched a show in which a puppet named Maxi put away a piece of chocolate in a box before leaving the room. While Maxi was out, his mother found the chocolate and moved it to a cupboard. Maxi returns to the scene, the show is stopped, and children are asked to predict where Maxi will go to look for his chocolate. If the child says Maxi will look in the cupboard, she shows that she doesn't have a theory of mind. If the child says Maxi will look in the box, she passes, and is thought to have a theory of mind because the child demonstrates that she can attribute mental states and use them to predict Maxi's behavior.

The theory of mind research program was closely associated with a debate on folk psychology between folk psychology as theory (the standard view that human knowledge of other minds is theoretical in nature), and folk psychology as simulation (the view that our knowledge of other minds relies on using our own mind as a model). In the late 90's there was a growing acceptance that both theory-theory and simulation theory were partially right and partially wrong, which culminated in a general acceptance of some sort of hybrid theory (e.g. Nichols & Stich 2003; Goldman 2006). These arguments make use of empirical data from both the developmental and the animal cognition literature.

During this time, there were various attempts to uncover theory of mind in animals using nonverbal paradigms, without much success (Heyes 1998). Given the subsequent theoretical and definitional disagreements, some researchers have concluded that “the generic label ‘theory of mind’ actually covers a wide range of processes of social cognition” (Tomasello et al. 2003b, 239). The theory of mind research paradigm in animal cognition subsequently shifted from attempts to come up with a nonverbal false belief task, toward more specific questions about cognitive capacities like understanding others' perceptual states (Hare et al. 2000), goals (Uller 2004), or intentionality (Tomasello 2005).

Like belief states, other's perceptual states are opaque, and require the attributor to make a distinction between oneself and another. And like knowing someone's beliefs states, knowing another's perceptual state can lead to predictions about future behavior. Ethological evidence that chimpanzees monitor gaze and modify their behavior when they are visible to others (e.g. Plooij 1978; Whiten & Byrne 1988; Goodall 1986) was taken as some evidence that chimpanzees can attribute perceptual states to others, and experimental researchers decided to design studies to determine whether chimpanzees understand seeing.

The results of early laboratory studies were mixed; David Premack's research suggested that chimpanzees do understand seeing (reviewed in Premack & Premack 2003), whereas studies by Povinelli and Eddy (1996) challenged that conclusion. Later studies suggested that chimpanzees understand both seeing and intentionality (Hare et al. 2000; Hare et al. 2001). In Hare et al.‘s experimental set-up, a subordinate and a dominant chimpanzee are released in a room baited with food. Normally, if both animals can see the food, or saw one another witness the baiting, a subordinate animal will avoid the food and allow the dominant access. However, in these experiments, when the food is occluded from the dominant's view, the subordinate will approach it. Only if the dominant can see the food or the baiting will the subordinate avoid it. The animals are across the room from one another, so the subordinate has to consider the visual perspective of the dominant in order to judge correctly whether he can see the food or not. Because it seems that the subordinate is able to make different judgments about whether to seek out the food based only on whether it is visible to the dominant, this study is thought to indicate that the apes understand the mental state of seeing.

Povinelli and Vonk (2004) criticize the Hare et al. studies, suggesting that the ecological nature of the study (using food competition behavior from the subject's natural repertoire) is a weakness of the study, not a strength as Hare et al. believed. Povinelli and Vonk argue that the subordinate chimpanzee need not have any understanding of the dominant's mental state, but could instead be operating on a theory of behavior. In response, the authors claim that they have accounted for all possible alternative explanations for the subordinate's behavior, making an inference to the best explanation argument that the subordinate understands what the dominant sees (Tomasello et al. 2003a, 2003b). Further evidence is provided by a study of chimpanzee hiding behavior (Hare et al. 2006). The debate is an epistemological one regarding how best to determine from evidence whether an animal postulates a mental state (Andrews 2005).

While the research on understanding what conspecifics can see has focused on chimpanzees and to a lesser extent other great apes, there are a few studies on other species. Research on scrub-jay caching behavior shows that individuals who have pilfered another's cache in the past will privately recache food when a conspecific observed the original caching, but not if the original caching was unobserved (Emery & Clayton 2004). Naive scrubjays did not recache. Emery & Clayton suggest that the jays who do recache are engaging in what they call “experience projection. . .they relate information about their previous experience as a pilferer to the possibility of future stealing by another individual, and modify their recovery strategy appropriately” (Emery & Clayton 2004, 1905). Experience projection, they suggest, relates to theory of mind.

Studies on perceptual understanding were also done with rhesus monkeys (Flombaum & Santos 2005). Like the chimpanzee and the scrubjay studies, these experiments set up a naturalistic competitive situation in which the subject had to predict the behavior of a competitor. In one version of this study, rhesus macaques from the island of Cayo Santiago were pitted against human competitors in a foraging task; two experimenters would approach a lone monkey, and each would situate himself differently so that the monkey was visible to one experimenter but not the other. Both experimenters had one grape. Flombaum and Santos found that monkeys were more likely to steal grapes from the experimenter who couldn't see them. They found similar results for audibility; when given the choice of stealing a grape in a transparent box covered with bells, or a grape in a transparent box that was free of noisemakers, the monkeys preferred the silent food when no one was looking at them. However, when it was obvious that the monkey was observed, there was no preference for stealing quiet over noisy grapes (Santos et al. 2006).

Other studies in this area focus on an understanding of intentionality and goal directed behavior. For example, Claudia Uller used a dishabituation paradigm designed to test for understanding of goal directed behavior in human infants to test infant chimpanzees (Uller 2004). Recall that dishabituation paradigms involve showing the subject a stimulus until it no longer holds the subject's attention, and then showing them one of two alternative stimuli. If one of the new stimuli gets more attention than the other, it is thought that the subject sees that stimulus as different from the original target stimulus. Gergely and colleagues used this method to study goal attribution in human infants: 12-month-old infants were habituated to a video of a small ball leaping over a barrier to reach a large ball (Gergely et al. 1995). After habituation, the infants were shown either the identical behavior with the barrier removed (e.g., the small ball moving toward the big ball and then jumping in the air before meeting the large ball) or a different behavior in which the small ball moved directly toward the large ball. Children looked longer at the first condition, and the authors concluded that infants perceive the small ball's action as goal-directed. Using the same paradigm, Uller found that chimpanzee infants respond at the same rate as human infants, but she suggests more evidence is required before claiming that chimpanzees understand intentions (Uller 2004).

Another recent study of understanding intentionality also compared the behavior of chimpanzees and human children (Warneken & Tomasello 2006). In a naturalistic social setting, subjects were nonverbally requested to help the experimenter achieve goals, such as picking up a dropped sponge or opening a box. They found that 18 month-old children and chimpanzees both respond to simple requests (e.g. picking up a dropped object), but that children are also able to respond to more complex requests. Earlier studies by Call et al. also demonstrated that chimpanzees are able to distinguish between a person who is unwilling to perform a task, and one who is unable (Call et al. 2004).

While chimpanzees show some sensitivity to intentions and goals, domestic dogs may be even more attuned to the intentions of humans. Dogs are able to use the gaze of a human in order to determine where food is hidden, an ability not demonstrated in the chimpanzee (Hare et al. 1998; Hare & Tomasello 1999; Miklosi & Topal 2004; Brauer et al. 2006). Dogs appear to be sensitive to eye gaze in humans, and often make eye contact before initiating play. One explanation for dogs' social acuity is that in selecting for traits that make dogs better human companions, humans inadvertently bred dogs who are better able to pass theory of mind tasks (Hare et al. 2002).

4.2.2 Metacognition

Research on metacognition aims to explore what individuals know about their own minds. One area of much attention has been mirror self-recognition (MSR). In this paradigm, developed by psychologist Gordon Gallup, subjects are surreptitiously marked and then given a mirror. “Passing” MSR involves touching the mark more frequently when there is a mirror available than when there is not. Gallop theorized that passing MSR entails that the animal has a concept of self (Gallup 1970), though others dispute this claim. While it was once thought to be a rare behavior, limited to some of the great apes, today many species have been studied and at least some positive results have been reported for the following species:

Many other species failed to show mirror directed behavior, including some monkey species, which suggests to some that there is a corresponding cognitive mechanism that the above species, but not others, enjoy. However, it has been pointed out that there are ecological and biological constraints on this test; not all species are visually oriented, and some find eyes aversive (this was the explanation for studies that failed to show MSR in gorillas). For a discussion of these issues, see the collection of articles in Self-awareness in Animals and Humans (Parker et al. 1994).

Uncertainty monitoring is another area of research aimed to investigate understanding of one's own mental state (for a review, see Shettleworth & Sutton 2006). Subjects who know what they do and do not know demonstrate metacognition about their epistemic states, and several nonverbal tests for uncertainty monitoring have been developed for use with different species. The paradigm might go as follows: subjects are trained to indicate whether a stimulus is the same as or different from a sample. When the subjects get the answer correct, they are rewarded with food, but food is taken away when they get the wrong answer. Once the subjects are trained on this task, the paradigm is modified to introduce a “bail out” key with the function of starting a new trial without supplying either a reward or punishment. Interspersed with the easy stimuli are ambiguous stimuli that the subject is unable to accurately categorize above a chance level. If the subjects learn to chose the “bail out” key when they are uncertain, it is thought to indicate that the subjects are aware of their epistemic state. It has been reported that many species choose the “bail out” key in such a way as to maximize rewards, including dolphins (Smithe et al. 1995), rhesus monkeys (Hampton 2001), great apes (Call & Carpenter 2001), human infants (Call & Carpenter 2001). Mixed results have been reported with pigeons (Sole et al. 2003).

Though such tests have been designed to test for metacognition, Peter Carruthers argues that animals can come to solve the problems without engaing in second-order reasoning. He suggests that the animal could be operating over beliefs and desires of different strengths, and that standard practical reasoning systems can be used to output different responses to the different permutations of weak and strong beliefs and desires (Carruthers forthcoming).

4.3.1 Emotions

Social cognition isn't limited to knowing the reasons another has for acting; it also involves understanding the emotions of others. While out of vogue for some time, researchers are again attending to emotions in different animal species, and the ability of some species to perceive the emotions of others. Studying emotion in other species is an important part of studying cognition because knowledge of another's emotional state allows an actor to respond differently in otherwise similar situations. Among social species, awareness of others' emotions is thought to play an important role in regulating social interactions, coordinating behavior, forming bonds between mothers and infants, as well as in forming short-term coalitions and long lasting relationships.

At least since Darwin's The Expression of the Emotions in Man and Animals , facial expressions have been of interest because they can indicate individuals' affective states. Research from the field and the lab suggests that different facial expressions have the same kinds of functions for chimpanzees and humans (e.g. van Hoof 1967, 1972; Goodall 1986; Parr 2001). Just as Paul Ekman argued for universality in emotional expressions among humans across cultures (Ekman et al. 1969), animal researchers have argued that at least some human emotions are also found in chimpanzees, and that chimpanzee facial expressions are homologous to human facial expressions in morphology and function. For example, van Hoof has argued that the bare-teeth display of chimpanzees is homologous to the human smile (van Hoof 1973).

Lisa Parr's research demonstrates that chimpanzees, like human infants, are able to categorize facial expressions associated with different emotional responses. Using a match-to-sample paradigm, Parr and colleagues have shown that chimpanzees recognize at least five different facial expressions: bared-teeth display, scream, pant-hoot, play face, and relaxed-lip face (Parr et al. 1998). Given the salience of these facial expressions to chimpanzees, Parr and colleagues have argued that facial expressions are important behaviors for regulating social relations (Parr & de Waal 1999; Parr et al. 2000). Current research in Parr's lab is focused on the development of ChimpFACS (Facial Action Coding System — see Other Internet Resoures) modeled after Ekman's work in emotion in human facial expressions. They are using the ChimpFACS to construct models of chimpanzee expressions in order to determine the configuration of muscle movements that the chimpanzees find salient in their perception of emotion.

Another area of study in animal emotion focuses on stress. As a response to potentially dangerous situations, stress is thought to be an adaptive emotion in the short term for humans and other animals. Stress is measured physiologically via levels of glucocorticoid hormones such as cortisol. In humans cortisol levels are correlated with stress levels, and researchers have studied stressors such as dominance and status ranking in a number of different species (Sapolsky 2005; Abbott et al. 2003). For example, among baboons stress in the form of elevated glucocorticoid levels has been documented in females for about a month after the death of a close relative, in nursing mothers when a potentially infanticidal immigrant male arrives, in females when the female ranking system is undergoing instability, and in males when the male ranking system is unstable (Cheney & Seyfarth 2007). However, baboons do not appear to experience stress in the face of another's stress. Females are not concerned about male rank instability, and the death of a cohort's infant does not raise stress levels in anyone but the mother. It is suggested that for the baboon, stress is personal and egocentric, and the lack of sensitivity to the stress of others might be indicative of a lack of empathy as well (Cheney & Seyfarth 2007).

4.3.2 Empathy and Morality

Some research in moral psychology suggests that empathy is a necessary component of moral agency, and animal cognition researchers have been examining whether any other species share this ability with humans. Empathy is thought to require the same cognitive sophistication as does understanding another's mental state or intention, but in addition it requires an affective response to that mental state. While the terms ‘empathy’ (sharing the mental state of another) and ‘sympathy’ (a friendly feeling in response to another's mental state) have distinct meanings in the history of psychology, folk psychology, and ethical theory, in the animal cognition research the senses are often blurred and the terms used interchangeably.

Early reports of chimpanzee empathy came from Russian comparative psychologist N. N. Ladygina-Kohts, who raised a chimpanzee named Joni in the early 20 th century (Ladygina-Kohts 2002). More recently, Sanjida O'Connell analyzed thousands of qualitative reports of primate responses to the distress of others, and her results suggest that apes give complex responses in the face of others' emotions, compared to the responses of monkeys in similar situations (O'Connell 1995). Studies of chimpanzee behavior performed by Frans de Waal and his colleagues suggest that chimpanzees understand emotions, and respond to different emotional states with different behavior, e.g. consoling the looser of a fight, helping, etc. De Waal takes these behaviors to be evidence of empathy in chimpanzees (de Waal 2006).

Much of the research on empathy in other species examines helping behavior. Helping is ubiquitous among humans, even when it requires that the actor suffer a cost, and even when the recipient is a stranger. But because helping can be engaged in without empathy, and empathic helping requires additional cognitive resources (e.g. knowledge about how to help someone achieve their goal and the motivation to act on that knowledge), it is important to understand the limitations of these studies.

Reports of naturalistic animal behavior suggest that many nonhuman animal species engage in prosocial behavior that may be empathic or proto-empathic in nature (de Waal 1996). De Waal often presents the famous example of Binti-Jua, a female gorilla at the Brookfield zoo in Chicago, who made the news when she rescued a 3 year-old human boy who fell into her habitat. Binti-Jua cradled the boy in her arms before handing him to a zookeeper. However, critics of the prosocial interpretation of Binti-Jua's behavior suggest that given her early exposure to doll play, an associative learning explanation is also possible.

Others claim that what may look like prosocial behavior may instead be a way of eliminating aversive stimuli. For example, research on rats and rhesus monkeys has shown that both species will cease eating when doing so causes shocks to a conspecific in an adjoining cage (Masserman et al. 1964). Masserman reports that one rhesus monkey almost starved himself to death to avoid shocking another. Alternatively, helping behavior among kin may be explained noncognitively as biological altruism . To determine whether other species engage in helping behavior that cannot be explained by other mechanisms, researchers have developed paradigms to determine whether chimpanzees display helping behavior to unrelated individuals. Chimpanzees are thought to be an especially good species to examine, given the range of cooperative behaviors they naturally perform, such as hunting (Boesch 2002), border patrolling (Mitani 2002), and coalition building (de Wall 1982). Cooperation among chimpanzees (Hirata 2003; Melis et al. 2006) and bonobos (Hare et al. 2007) has been demonstrated in a food-sharing task, but chimpanzees are thought to cooperate only when the dyads are generally tolerant of one another (Hare et al. 2007).

Another set of experiments suggests that chimpanzees are indifferent to the desires of others, and will not help another access food even when it requires no additional effort on the actor's part (Silk et al. 2005). The experimental setup allowed the actor to pull one of two ropes, one of which delivered food to only the actor, and the other of which delivered food to an adjacent chimpanzee as well as the actor. No preference was found for pulling the rope that rewarded both animals, even though the actor could see and hear the other chimpanzee. There was no significant correlation between another animal being present and which rope was pulled. Silk and colleagues conclude that “The absence of other-regarding preferences in chimpanzees may indicate that such preferences are a derived property of the human species, tied to sophisticated capacities for cultural learning, theory of mind, perspective taking and moral judgment” (Silk et al. 2005, 1359).

In contrast to Silk and colleagues' findings, several studies coming from the Max Planck Institute indicate that chimpanzees may be willing to help others. In one set of studies Warneken and Tomasello compared the helping behavior of 18 month-old human infants and chimpanzees, and found that while both subjects helped an adult human male retrieve dropped or out of reach objects, the infants were more likely to help in other contexts as well. The authors explain the disparity between Silk et al.‘s results and their own by suggesting that the chimpanzees are distracted by the presence of food; they suggest other-regarding preferences in chimpanzees are overridden by the opportunity to gain food for themselves (Warneken & Tomasello 2006). A second study led by Warneken demonstrated that when a chimpanzee wasn't able to receive food for himself, he would reliably help another chimpanzee acquire food by opening a door (Warneken et al. 2007).

Insofar as these studies indicate empathy in a chimpanzee, one would expect to find a number of related cognitive capacities, including a sophisticated social cognition that allows for the reading of goals or emotions, and a causal reasoning capacity that allows the animal to offer assistance. Whether the current findings demonstrate that chimpanzees display altruisim, as is sometimes claimed (de Waal 2006), or whether having a full fledged theory of mind is a necessary part of our moral psychology (Hauser 2005), are issues that require conceptual analysis as well as empirical investigation of prosocial behavior in humans and other animals.

Today we are in a kind of golden era when it comes to animal cognition research. Different species are being studied in the field and in the lab, and the results of these studies may be relevant to areas of philosophy including action theory, agency, belief, concepts, consciousness, culture, epistemology, ethics, folk psychology, imagery, language, memory, mental causation, mental content, modularity of mind, perception, personal identity, practical reason, rationality, and so forth. It seems that every day a new report is released, and many of these may have some theoretical implications.

Of course, scientific reports must be examined carefully to distinguish between methodologically solid findings and unwarranted interpretations, and it goes without saying that popular media reports of these studies are sometimes misleading. The epistemology of animal cognition research is ripe for investigation. Animal cognition is studied by psychologists, anthropologists, biologists, zoologists, neuroscientists, and ecologists, among others, and while some textbooks aim to integrate these different disciplinary approaches (e.g. Shettleworth 1998), and many extremely clever research methods have been used to test hypotheses, there is no unifying principle that brings together all those working under the umbrella of animal cognition. For example, psychologists tend to endorse the old adage: “It's not what you do, it's how you do it” to describe their interest in mechanism, while the ecologists' focus on adaptive problems may be less interested in mechanism, since solving adaptive problems such as acquiring food and avoiding predators might involve several cognitive processes.

Philosophy of animal cognition as sub-field of philosophy of science is one place where these methodological questions can be examined. For further reading in this area, see Colin Allen and Marc Bekoff's classic book Species of Mind: The Philosophy and Biology of Cognitive Ethology . For an introduction to an evolutionary psychology approach to studying animal cognition, see another classic, Sara Shettleworth's Cognition, Evolution, and Behaivor . And one more classic, from ethology, is Philip Lehner's Handbook of Ethological Methods .

• Abbott, D. H., Keverne, E. B., Bercovitch, F. B., Shively, C. A., Mendoza, S. P., Saltzman, W., Snowdon, C. T., Ziegler, T. E., Banjevic, M., Garland, T., Jr., & Sapolsky, R. M. (2003). “Are Subordinates Always Stressed? A Comparative Analysis of Rank Differences in Cortisol Levels Among Primates.” Hormones and Behavior, 43: 67-82.
• Allen, C. & Hauser, M. (1996). “Concept Attribution in Nonhuman Animals: Theoretical and Methodological Problems in Ascribing Complex Mental Processes.” In M. Bekoff & D. Jamieson (Eds.), Readings in Animal Cognition . Cambridge MA, MIT Press. 47-62.
• Allen, C. & Bekoff, M. (1997). Species of Mind: The Philosophy and Biology of Cognitive Ethology . Cambridge MA, MIT Press.
• Allen, C. & Saidel, E. (1998). “The Evolution of Reference.” In D. Cummins & C. Allen (Eds.), The Evolution of Mind . New York, Oxford University Press. 183-202.
• Allen, C. (1999). “Animal Concepts Revisited: The Use of Self-Monitoring as an Empirical Approach.” Erkenntnis, 51(1): 33-40.
• Allen, C. (2004). “Is Anyone a Cognitive Ethologist?” Biology and Philosophy, 19: 589-607.
• Allen-Hermanson, S. (2005). “Morgan's Canon Revisited.” Philosophy of Science, 72: 608-631.
• Andrews, K. (2002). “Interpreting Autism: A Critique of Davidson on Thought and Language.” Philosophical Psychology , 15: 317-332.
• Andrews, K. (2005). “Chimpanzee Theory of Mind: Looking in all the Wrong Places?” Mind and Language, 20: 521-536.
• Andrews, K. (forthcoming). “Politics or Metaphysics? On Attributing Psychological Properties to Animals.” Biology and Philosophy .
• Aquinas, S. T. (1981). Summa Theologica . Grand Rapids, Christian Classics.
• Aristotle. (1999). The Metaphysics . New York, Penguin Classics.
• Asquith, P. J. (1991). “Primate Research Groups in Japan: Orientations and East-West Differences.” In L. Fedigan & P. J. Asquith (Eds.), The Monkeys of Arashiyama. Thirty-five Years of Research in Japan and the West (pp. 81-98). Albany: SUNY Press.
• Asquith, P. J. (1997). “Why Anthropomorphism is not Metaphor: Crossing Concepts and Cultures in Animal Behavior Studies.” In R. W. Mitchell, N. S. Thompson & H. L. Miles (Eds.), Anthropomorphism, Anecdotes, and Animals (pp. 22-36). Albany: State University of New York.
• Baillargeon, R. & DeVos, J. (1991). “Object Permanence in Young Infants: Further Evidence.” Child Development, 62: 1227-1246.
• Barlow, H. (1990). “The Mechanical Mind.” Annual Review of Neuroscience, 13: 15-24.
• Barrett, L., & Henzi, P. (2005). “The Social Nature of Primate Cognition.” Proceedings of the Royal Society B: Biological Sciences, 272: 1865-1875.
• Barton, R., & Dunbar, R. I. M. (1997). “Evolution of the Social Brain.” In A. Whiten & R. Byrne (Eds.), Machiavellian Intelligence II: Evaluations and Extensions (pp. 240-263). Cambridge: Cambridge University Press.
• Beck, B. B. (1982). “Chimpocentrism: Bias in Cognitive Ethology.” Journal of Human Evolution, 11: 3-17.
• Beecher, M. D., Burt, J. M., O'Loghlen, A. L., Templeton, C. N., & Campbell, S. E. (2007). “Bird Song Learning in an Eavesdropping Context.” Animal Behavior, 73: 929-935.
• Bekoff, M., & Allen, C. (1997). “Cognitive Ethology: Slayers, Skeptics and Proponents.” In R. W. Mitchell, N. S. Thompson & H. L. Miles (Eds.), Anthropomorphism, Anecdotes, and Animals (pp. 313-334). Albany: State University of New York Press. 313-334.
• Bekoff, M. (2001). “Social Play Behaviour: Cooperation, Fairness, Trust, and the Evolution of Morality.” Journal of Consciousness Studies, 8: 81-90.
• Bennett, J. (1978). “Some Remarks about Concepts.” Behavioral and Brain Sciences, 1: 557-560.
• Bermúdez, J. (2003). Thinking Without Words . Cambridge MA, MIT Press.
• Bermúdez, J. (2006). “Animal Reasoning and Proto-Logic.” In S. Hurley and M. Nudds (Eds.), Rational Animals? Oxford, Oxford University Press. 127-138.
• Beynon, P., & Rasa, O. A. E. (1989). “Do Dwarf Mongooses Have a Language? Warning Vocalizations Transmit Complex Information.” Suid-Afrikaanse Tydskrif vir Wetenskap, 85: 447-450.
• Blumberg, M. S., & Wasserman, E. A. (1995). “Animal Mind and the Argument from Design.” American Psychologist, 50: 133-144.
• Boesch, C. (2002). “Cooperative Hunting Roles among Taie Chimpanzees.” Human Nature, 13: 27-46.
• Boysen, S. T. & Berntson, G. G. (1989). “Numerical Competence in a Chimpanzee.” Journal of Comparative Psychology, 103: 23-31.
• Brauer, J., Kaminski, J., Riedel, J., Call, J., & Tomasello, M. (2006). “Making Inference about the Location of Hidden Foods: Social Dog, Causal Ape.” Journal of Comparative Psychology, 120: 38-47.
• Brown, J. L. (1975). The Evolution of Behavior . New York: W.W. Norton.
• Byrne, R. & Whiten, A. (Eds.) (1988). Machiavellian Intelligence: Social Expertise and the Evolution of Intellect in Monkeys, Apes, and Humans . New York: Oxford University Press.
• Byrne, R. (1997). “What's the Use of Anecdotes? Distinguishing Psychological Mechanisms in Primate Tactical Deception.” In R. W. Mitchell & N. S. Thompson & H. L. Miles (Eds.), Anthropomorphism, Anecdotes, and Animals . Albany: SUNY Press.
• Byrne, R. & Russon, A. (1998). “Learning by Imitation: A Hierarchical Approach.” Behavioral and Brain Sciences, 21: 667-721.
• Byrne, R. (1999). “Imitation Without Intentionality: Using String Parsing to Copy the Organization of Behavior.” Animal Cognition, 2: 63-72.
• Call, J. & Carpenter, M. (2001). “Do Apes and Children Know What They Have Seen?” Animal Cognition , 4: 207-220.
• Call J., Hare B.H., Carpenter M. & Tomasello M. (2004). “‘Unwilling’ versus ‘Unable’: Chimpanzees' Understanding of Intentional Action.” Developmental Science , 7(4): 488-498.
• Candland, D.K. (1993). Feral Children and Clever Animals . New York: Oxford University Press.
• Carruthers, P. (forthcoming). “Meta-cognition in Animals: A Skeptical Look.” Mind and Language .
• Cavalieri, P., & Singer, P. (1994). The Great Ape Project: Equality Beyond Humanity . New York: St. Martin's Press.
• Chater, N. & Heyes, C. (1994). “Animal Concepts: Content and Discontent.” Mind and Language, 9: 209-246.
• Cheney, D.L. & Seyfarth, R.M. (1990). How Monkeys See the World: Inside the Mind of Another Species . The University of Chicago Press.
• Cheney, D.L. & Seyfarth, R.M. (2007). Baboon Metaphysics: The Evolution of a Social Mind . University Of Chicago Press.
• Chomsky, N. (1968). Language and Mind . New York, Harcourt, Brace and World.
• Chomsky, N. (1980). “Human Language and Other Semiotic Systems”. In T. A. Sebeok & J. Umiker-Sebeok (Eds.), Speaking of Apes: A Critical Anthology of Two-Way Communication with Man . pp. 429-440. New York: Plenum Press.
• Cosmides, L. & Tooby, J. (1994). “Origins of Domain-Specificity: The Evolution of Functional Organization.” In L. Hirschfeld & S. Gelman (Eds.), Mapping the Mind: Domain-Specificity in Cognition and Culture . New York: Cambridge University Press. 85-116.
• Crist, E. (1999). Images of Animals: Anthropomorphism and Animal Mind . Philadelphia: Temple University Press.
• Darwin, C. (1974). The Descent of Man and Selection in Relation to Sex . Detroit: Gale Research.
• Daston, L., & Mitman, G. (Eds.). (2005). Thinking with Animals: New Perspectives on Anthropomorphism . New York: Columbia University Press.
• Davidson, D. (1975). “Thought and Talk.” In S. Guttenplan (Ed.), Mind and Language . Oxford: Oxford University Press. 7-23.
• Davidson, D. (1982). “Rational Animals.” Dialectica, 36: 317-328.
• Dennett, D. (1978). “Beliefs about Beliefs.” Behavioral and Brain Sciences, 4: 568-570.
• Dennett, D. (1983). “Intentional Systems in Cognitive Ethology: The ‘Panglossian Paradigm’ Defended.” Behavioral and Brian Sciences, 6: 343-390.
• Dennett, D. (1987). The Intentional Stance . Cambridge, MA: MIT Press.
• Dennett, D. (1995). “Do Animals Have Beliefs?” In H. Roitblat & J. Meyer (Eds.), Comparative Approaches to Cognitive Science : MIT Press. 111-118.
• Descartes, R. (2006). A Discourse on the Method . Oxford, Oxford University Press.
• Dretske, F. (1988). Explaining Behavior: Reasons in a World of Causes . Cambridge: MIT Press
• Dretske, F. (2006). “Minimal Rationality.” In S. Hurley & M. Nudds (Eds.), Rational Animals? pp. 107-116. Oxford: Oxford University Press.
• Dunbar, R. (1992). “Neocortex Size as a Constraint on Group Size in Primates.” Journal of Human Evolution, 20: 469-493.
• Dunbar, R. (1998). “The Social Brain Hypothesis.” Evolutionary Anthropology, 6: 178-190.
• Ekman, P., Sorenson, E. R., & Friesen, W. V. (1969). “Pan-Cultural Elements in Facial Displays of Emotions.” Science 164(3875): 86-88.
• Emery, N. (2000). “The Eyes Have It: The Neuroethology, Function, and Evolution of Social Gaze.” Neuroscience and Biobehavioral Reviews, 24: 581-604.
• Emery, N.J. & Clayton, N.S. (2004). “Western Scrub-Jays (Aphelocoma Californica) Use Cognitive Strategies to Protect Their Caches from Thieving Conspecifics.” Animal Cognition 7: 37-43.
• Evans, C. & Marler, P. (1995). “Language and Animal Communication: Parallels and Contrasts.” In H. L. Roitblat & J. A. Meyer (Eds.), Comparative Approaches to Cognitive Science . Cambridge MA: MIT Press. 341-382.
• Evans, C. (1997). “Referential Signals.” In D. Owings, M. D. Beecher & N. S. Thompson (Eds.), Perspectives in Ethology, Vol.12: Communication . New York: Plenum. 99-143.
• Evans, C.S., Evans, L. & Marler, P. (1993a). “On the Meaning of Alarm Calls: Functional Reference in an Avian Vocal System.” Animal Behaviour , 46: 23-38.
• Evans, C. S., Macedonia, J. M. & Marler, P. (1993b). “Effects of Apparent Size and Speed on the Response of Chickens (Gallus Gallus) to Computer-Generated Simulations of Aerial Predators.” Animal Behaviour , 46: 1-11.
• Evans, C., & Evans, L. (1999). “Chicken Calls are Functionally Referential.” Animal Behavior, 58: 307-319.
• Evans, C. S. (2002). “Cracking the Code: Communication and Cognition in Birds.” In M. Bekoff, C. Allen & G. M. Burghardt (Eds.), The Cognitive Animal: Empirical and Theoretical Perspectives on Animal Cognition . Cambridge, MA: MIT Press. 315-322.
• Evans, C. S., & Evans, L. (2007). “Representational Signaling in Birds.” Biology Letters, 3: 8-11.
• Fisher, J. A. (1990). “The Myth of Anthropomorphism.” In M. Bekoff & D. Jamieson (Eds.), Interpretation and Explanation in The Study Of Animal Behavior: Vol. 1, Interpretation, Intentionality, and Communication . Boulder: Westview Press. 96-116.
• Fisher, J. A. (1991). “Disambiguating Anthropomorphism: An Interdisciplinary Review.” In P. P. G. Bateson & P. H. Klopfer (Eds.), Perspectives in Ethology (Vol. 9, pp. 49-85). New York: Plenum.
• Flombaum, J. I., & Santos, L., R. (2005). “Rhesus Monkeys Attribute Perceptions to Others.” Current Biology, 15: 447-452.
• Fouts, R.S., D.H. Fouts, et al. (1989). “The Infant Loulis Learns Signs from Cross-Fostered Chimpanzees.” Teaching sign Language to Chimpanzees . A.R. Gardner, B.T. Gardner and T. Van Cantfort. Albany, State University of New York Press: 280-292.
• Gallup, G.G. (1970). “Chimpanzees: Self-Recognition.” Science, 167: 341-343.
• Gardner, B.T. & Gardner, A.R. (1971). “Two-Way Communication with an Infant Chimpanzee.” Behavior of Nonhuman Primates . A. M. Schrier and F. Stollnitz. New York, Academic Press: 117-184.
• Gardner, R.A. & Gardner, B.T. (1978). “Comparative Psychology and Language Acquisition.” Annals of the New York Academy of Sciences , 309: 37-76.
• Gergely, G., Nadasdy, Z., et al. (1995). “Taking the Intentional Stance at 12 Months of Age.” Cognition 56: 165-193.
• Golani, I. (1992). “A Mobility Gradient in the Organization of Movement: The Perception of Movement through Symbolic Language.” Behavioral and Brain Sciences, 15: 249-308.
• Goldman, A.I. (2006). Simulating Minds: The Philosophy, Psychology, and Neuroscience of Mindreading . Oxford: Oxford University Press.
• Gómez, J. C. (2005). “Species Comparative Studies and Cognitive Development.” Trends in Cognitive Science, 9(3): 118-125.
• Goodall, J. (1986). The Chimpanzees of Gombe: Patterns of Behavior . Cambridge, MA: Harvard University Press.
• Greene, E., & Meagher, T. (1998). “Red Squirrels, Tamiasciurus Hudsonicus, Produce Predator-Specific Alarm Calls.” Animal Behavior, 55: 511-518.
• Greenfield, P. (1991). “Language, Tools, and Brain: The Ontogeny and Phylogeny of Hierarchically Organized Sequential Behavior.” Behavioral and Brain Sciences, 14: 531-595.
• Grice, H. P. (1969). “Utterer's Meaning and Intention.” The Philosophical Review, 78: 147-177.
• Griffin, D. R. (1984). Animal Thinking , Harvard University Press.
• Griffin, D. R. (1985). “Animal Consciousness.” Neuroscience and Biobehavioral Reviews, 9: 615-22.
• Griffin, D. R. (1992). Animal Minds . Chicago: University of Chicago Press.
• Haesler S, Rochefort C, Georgi B, Licznerski P, Osten P, & Scharff, C. (2007). “Incomplete and Inaccurate Vocal Imitation after Knockdown of FoxP2 in Songbird Basal Ganglia Nucleus Area X.” PLoS Biology 5(12): 2885-2897.
• Hampton, R.R. (2001). “Rhesus Monkeys Know When They Remember.” Proceedings of the National Academy of Sciences of the United States of America , 98: 5359-5362.
• Hare, B., Call, J., & Tomasello, M. (1998). “Communication of Food Location Between Human and Dog (Canis Familiaris).” Evolution of Communication, 2: 137-159.
• Hare, B. & Tomasello, M. (1999). “Domestic Dogs (Canis Familiaris) use Human and Conspecific Social Cues to Locate Hidden Food.” Journal of Comparative Psychology, 113: 1-5.
• Hare, B., Call, J., Agnetta, B., & Tomasello, M. (2000). “Chimpanzees Know what Conspecifics Do and Do Not See.” Animal Behavior, 59: 771-785.
• Hare, B., Call, J., & Tomasello, M. (2001). “Do Chimpanzees Know What Conspecifics Know?” Animal Behavior, 6(1), 139-151.
• Hare, B,, Brown, M., Williamson, C. & Tomasello, M. (2002). “The Domestication of Social Cognition in Dogs.” Science , 298: 1634-1636
• Hare, B., Call, J., & Tomasello, M. (2006). “Chimpanzees Deceive a Human by Hiding.” Cognition, 101: 495-514.
• Hare, B., Melis, A., Woods, V., Hastings, S. & Wrangham, R. W. (2007). “Tolerance Allows Bonobos to Outperform Chimpanzees on a Cooperative Task.” Current Biology, 17: 619-623.
• Hauser, M.D., Kralik, J., Botto, C., Garrett, M. & Oser, J. (1995). “Self-Recognition in Primates: Phylogeny and the Salience of Species-Typical Traits.” Proceeding of the National Academy of the Sciences , 92: 10811-10814.
• Hauser, M.D., MacNeilage, P. & Ware, M. (1996). “Numerical Representations in Primates.” Proceeding of the National Academy of the Sciences , 93: 1514-1517.
• Hauser, M.D. & Carey, S. (1998). “Building a Cognitive Creature from a Set of Primitives: Evolutionary and Developmental Insights.” In D.D. Cummins & C. Allen (Eds.), The Evolution of Mind . New York: Oxford University Press, pp. 51-106.
• Hauser, M.D., Chomsky, N. & Fitch, W.T. (2002). “The Faculty of Language: What Is It, Who Has It, And How Did It Evolve?” Science , 298: 1569-1579.
• Hauser, M.D. (2005). “Our Chimpanzee Mind.” Nature , 437: 60-63.
• Hauser, M. D. (2006). Moral Minds: How Nature Designed Our Universal Sense of Right and Wrong . New York: Ecco.
• Hayes, C. (1951). The Ape in Our House . New York, Harper.
• Hayes, K. J. & Hayes, C. (1952). “Imitation in a Home-Raised Chimpanzee.” Journal of Comparative Psychology, 45: 450-459.
• Herman, L.M. (1986). “Cognition and Language Competencies of Bottlenosed Dolphins.” In R. J. Schusterman, J. Thomas, and F. G. Wood (Eds.) Dolphin Cognition and Behavior: A Comparative Approach . pp. 221-251. Hillsdale, NJ: Lawrence Erlbaum Associates.
• Herrnstein, R. J., Loveland, D. H., & Cable, C. (1976). “Natural Concepts in Pigeons.” Journal of Experimental Psychology: Animal Behavior Processes, 2: 285-302.
• Herrnstein, R.J. (1979). “Acquisition, Generalization, and Discrimination Reversal of a Natural Concept .” Journal of Experimental Psychology: Animal Behavior Processes , 5: 116-129.
• Heyes, C. (1998). “Theory of Mind in Nonhuman Primates.” Behavioral and Brian Sciences, 21: 101-134.
• Hillix, W., & Rumbaugh, D. (2004). Animal Bodies, Human Minds: Ape, Dolphin, and Parrot Language Skills . New York: Kluwer.
• Hinde, R.A. (1970). Animal Behavior: A Synthesis of Ethology and Comparative Psychology (2nd ed.) . New York: McGraw-Hill.
• Hirata, S. (2003). “Cooperation in Chimpanzees.” Hattatsu, 95: 103-111.
• Hockett, C. (1977). The View From Language: Selected Essays. Athens, GA: University of Georgia Press.
• Holekamp, K. (2006). “Questioning the Social Intelligence Hypothesis.” Trends in Cognitive Science, 11: 65-69.
• Hollén, L. I., & Manser, M. B. (2007). “Motivation Before Meaning: Motivational Information Encoded in Meerkat Alarm Calls Develops Earlier than Referential Information.” The American Naturalist, 169: 758-767.
• Hopper, L., Spiteri, A., Lambeth, S. P., Schapiro, S. J., Horner, V., & Whiten, A. (2007). “Experimental Studies of Traditions and Underlying Transmission Processes in Chimpanzees.” Animal Behaviour, 73: 1021-1032.
• Horner, V., & Whiten, A. (2005). “Causal Knowledge and Imitation/Emulation Switching in Chimpanzess and Children.” Animal Cognition, 8: 164-181.
• Hume, D. (1968). A Treatise of Human Nature . Oxford, Oxford University Press.
• Humphrey, N. (1978). “Nature's Psychologists.” New Scientist , 900-904.
• Hurley, S. (2003a). “Making Sense of Animals: Interpretation vs. Architecture.” Mind and Language, 18: 231-256.
• Hurley, S. (2003b). “Animal Action in the Space of Reasons.” Mind and Language, 18: 231-256.
• Hurley, S., & Chater, N. (Eds.) (2005). Perspectives on Imitation From Neuroscience to Social Science- Volume 1: Mechaniams of Imitation and Imitation in Animals . Cambridge MA: MIT Press.
• Hurley, S. (2006). “Making Sense of Animals.” Rational Animals? S. Hurley and M. Nudds, Eds. Oxford, Oxford University Press: 139-174.
• Hurley, S. and M. Nudds, Eds. (2006). Rational Animals? Oxford, Oxford University Press.
• Inoue, S. & Matsuzawa, T. (2007). “Working Memory of Numerals in Chimpanzees.” Current Biology 17: R1004-R1005.
• Janik, V. M., Sayigh, L. S., & Wells, R. S. (2006). “Signature Whistle Shape Conveys Identity Information to Bottlenose Dolphins.” Proceedings of the National Academy of Sciences, 103: 8293-8297.
• Jolly, A. (1966). “Lemur Social Behavior and Primate Intelligence.” Science, 153: 501-506.
• Kant, I. (1963). “Duties to Animals and Spirits.” Lectures on Ethics . L. Infield. New York, Harper and Row: 239-241.
• Keeley, B. L. (2004). “Anthropomorphism, Primatomorphism, Mammalomorphism.” Philosophy and Biology, 19: 521-540.
• Keeton, W. T. (1967). Biological Science . New York: W.W. Norton.
• Kellogg, W. N. and L. A. Kellogg (1933). The Ape and the Child: A Comparative Study of the Environmental Influence Upon Early Behavior . New York, Hafner Publishing Company.
• Kennedy, J. S. (1992). The New Anthropomorphism . Cambridge: Cambridge University Press.
• King, B. (2004). The Dynamic Dance: Nonvocal Communication in African Great Apes . Cambridge, MA: Harvard University Press.
• Kiriazis, J. (C.N. Slobodchikoff). “Perceptual Specificity in the Alarm Calls of Gunnison's Prairie Dogs.” Behavioural Processes, 73: 29-35.
• Kohler, W. (1925). The Mentality of Apes . New York, Harcourt, Brace.
• Krutzen, M., Mann, J., Heithaus, M. R., Conner, R. C., Bejder, L., & Sherwin, W. B. (2005). “Cultural Transmission of Tool Use in Bottlenose Dolphins.” Proceedings of the National Academy of Sciences, 102: 8939-8943.
• Ladygina-Kohts, N.N. (2002). Infant Chimpanzee and Human Child: A Classic 1935 Comparative Study of Ape Emotions and Intelligence . New York: Oxford University Press.
• Lehner, P. N. (1996). Handbook of Ethological Methods (Second ed.). Cambridge: Cambridge University Press.
• Lin, A. C., Bard, K. A., & Anderson, J. R. (1992). “Development of Self-Recognition in Chimpanzees.” Journal of Comparative Psychology, 106: 120-127.
• Lloyd, E. (2004). “Kanzi, Evolution, and Language.” Biology and Philosophy, 19: 577-588.
• Locke, J. (2007). An Essay Concerning Human Understanding . Pomona, Italy: Pomona Press.
• Manser, M. B. (2001). “The Acoustic Structure of Suricates' Alarm Calls varies with Predator Type and the Level of Response Urgency.” Proceedings of the Royal Society B: Biological Sciences, 268: 2315-2324.
• Manser, M. B., Bell, M. B., & Fletcher, I. B. (2001). “The Information that Receivers Extract from Alarm Calls in Suricates.” Proceedings of the Royal Society B: Biological Sciences, 268: 2485-2491.
• Marino, L., Reiss, D., & Gallup, G. G. (1994). “Mirror Self-Recognition in Bottlenosed Dolphins: Implications for Comparative Investigations of Highly Dissimilar Species.” In S.T. Parker, R.W. Mitchell & M.L. Boccia (Eds.), Self-Awareness in Animals and Humans . Cambridge: Cambridge University Press, pp. 380-391.
• Marler, P. (1970). “Birdsong and Speech Development: Could there be Parallels?” American Scientist, 58: 669-673.
• Marler, P., Evans, C.S., & Hauser, M.D. (1992). “Animal Signals: Motivational, Referential, or Both?” In Papousek, H., Jürgens, U., and Papousek, M. (Eds.) Nonverbal Vocal Communication: Comparative and Developmental Approaches . Cambridge University Press, pp. 66-86.
• Masserman, J., Wechkin, M.S., & Terris, W. (1964). “Altruistic Behavior in Rhesus Monkeys.” American Journal of Psychiatry , 121: 584-85.
• Matsuzawa, T. (1985). “Use of Numbers by a Chimpanzee.” Nature 315: 57-59.
• Matsuzawa, T. (2003). “Koshima Monkeys and Bossou Chimpanzees: Culture in Nonhuman Primates Based on Long-Term Research.” In F. de Waal & P. Tyack (Eds.), Animal Social Complexity: Intelligence, Culture, and Individualized Societies . Cambridge, MA: Harvard University Press.
• Matsuzawa, T. (2006c). “Sociocognitive Development in Chimpanzees: A Synthesis of Laboratory Work and Fieldwork.” In T. Matsuzawa & M. Tomonaga & M. Tanaka (Eds.), Cognitive Development in Chimpanzees Tokyo: Springer. 3-36.
• Matsuzawa, T. (2006b). “Evolutionay Origins of the Human Mother-Infant Relationship.” In T. Matsuzawa & M. Tomonaga & M. Tanaka (Eds.), Cognitive Development in Chimpanzees Tokyo: Springer. 127-141.
• McGrew, W. C., & Tutin, C. E. G. (1978). “Evidence for a Social Custom in Wild Chimpanzees?” Man, 12: 234-251.
• McGrew, W. C., Marchant, L. F., Scott, S. E., & Tutin, C. E. G. (2001). “Intergroup Differences in a Social Custom of Wild Chimpanzees: The Grooming Hand-Clasp of the Mahale Mountains.” Current Anthropology, 42: 148-153.
• Melis, A.P., Hare, B., & Tomasello, M. (2006). “Chimpanzees Recruit The Best Collaborators.” Science , 311: 1297-1300.
• Menzel, E. W. J. (1969). “Naturalistic and Experimental Approaches to Primate Behavior.” In E. P. Willems & H. L. Rausch (Eds.), Naturalistic Viewpoints in Psychological Research (pp. 78-121). New York: Holt, Rinehart and Winston.
• Miklósi & Topal, J. (2004). “What Can Dogs Teach Us?” Animal Behavior , 67: 995-1004.
• Miles, H.L. (1983). “Apes and Language: The Search for Communicative Competence.” In J. de Luce and H. T. Wildner (Eds.) Language in Primates: Perspectives and Implications . New York: Springer-Verlag. 43-61.
• Millikan, R. G. (1993). White Queen Psychology and Other Essays for Alice . Cambridge, MA: MIT Press.
• Millikan, R. G. (2004). Varieties of Meaning: The Jean NIcod Lectures 2002 . Cambridge, MA, MIT Press.
• Millikan, R. G. (2006). “Styles of Rationality.” In S. Hurley and M. Nudds (Eds.), Rational Animals? Oxford, Oxford University Press. 117-126.
• Mitani, J. C., Watts, D. P., & Muller, M. N. (2002). “Recent Developments in the Study of Wild Chimpanzee Behavior.” Evolutionary Anthropology, 11: 9-25.
• Mitchell, R., Thompson, N., & Miles, H. L. (1997). Anthropomorphism, Anecdotes, and Animals . Albany: State Univeristy of New York Press.
• Mitchell, R. W. (1997). “Anthropomorphic Anecdotalism as Method.” In R. W. Mitchell & N. S. Thompson & H. L. Miles (Eds.), Anthropomorphism, Anecdotes, and Animals . Albany: SUNY Press. 151-169.
• Morgan, C. L. (1894). An Introduction to Comparative Psychology . London: W. Scott, limited.
• Nichols, S. and S. Stich (2003). Mindreading: An Integrated Account of Pretence, Self-Awareness, and Understanding Other Minds . Oxford, UK, Oxford University Press.
• O'Connell, S. M. (1995). “Empathy in Chimpanzees: Evidence for Theory of Mind?” Primates , 36(3): 397-410.
• Owings, D. H., & Hennessy, D. F. (1984). “The Importance of Variation in Sciurid Visual and Vocal Communication.” In J. O. Murie & G. R. Michener (Eds.), The Biology of Ground Dwelling Squirrels (pp. 171-200). Lincoln: University of Nebraska Press.
• Parker, S. T., Boccia, M. L., & Mitchell, R. W. (Eds.). (1994). Self-awareness in Animals and Humans: Developmental Perspectives. Cambridge: Cambridge University Press.
• Parr, L., Hopkins, W.D. & de Waal, F.B.M. (1998). “The Perception of Facial Expressions in Chimpanzees (Pan Troglodytes).” Evolution of Communication , 2: 1-23.
• Parr, L. A., & de Waal, F. (1999). “Visual Kin Recognition in Chimpanzees.” Nature, 399: 647-648.
• Parr, L. A., Winslow, J. T., Hopkins, W. D., & Waal, F. d. (2000). “Recognizing Facial Cues: Individual Recognition in Chimpanzees and Rhesus Monkeys.” Journal of Comparative Psychology, 114: 47-60.
• Parr, L. A. (2001). “Cognitive and Physiological Markers of Emotional Awareness in Chimpanzees (Pan Troglodytes).” Animal Cognition , 4: 223-229.
• Patterson, F. G. (1978). “The Gestures of a Gorilla: Language Acquisition in Another Pongid.” Brain and Language, 5: 72-97.
• Pearce, J. M. (1997). Animal Learning and Cognition . Hove: Psychology Press.
• Pepperberg, I. M. (1999). The Alex Studies: Cognitive and Communicative Abilities of Grey Parrots . Cambridge MA: Harvard University Press.
• Perry, S., Baker, M., Fedigan, L., Gros-Louis, J., Jack, K., MacKinnon, K. C., Manson, J. H., Panger, M., Pyle, K., & Rose, L. (2003). “Social Conventions in Wild White-Faced Capuchin Monkeys.” Current Anthropology, 44: 241-258.
• Perry, S. (2006). “What Cultural Primatology can Tell Anthropologists about the Evolution of Culture.” Annual Review of Anthropology, 35: 171-190.
• Petrides, M., Cadoret, G., & Mackey, S. (2005). “Orofacial Somatomotor Responses in the Macaque Monkey Homologue of Broca's Area.” Nature , 435(30), 1235-1238.
• Pfungst, O. (1965). Clever Hans (The Horse of Mr. Von Osten) . New York: Holt, Rinehart and Winston.
• Pinker, S., & Bloom, P. (1990). “Natural Language and Natural Selection.” Behavioral and Brain Sciences, 13: 707-784.
• Pinker, S. (1994). The Language Instinct . New York: W. Morrow and Co.
• Plooij, F.X. (1978). “Some Basic Traits of Language in Wild Chimpanzees?” In A. Lock (Ed.), Action, Gesture And Symbol (pp. 111-131). London: Academic Press.
• Plotnik, J. M., de Waal, F., & Reiss, D. (2006). “Self-Recognition in an Asian Elephant” Proceedings of the National Academy of Sciences, 103: 17053-17057.
• Povinelli, D. J. and T. J. Eddy (1996). “What Young Chimpanzees Know About Seeing.” Monographs of the Society for Research on Child Development 61(247): 1-189.
• Povinelli, D. J. (2000). Folk Physics for Apes . Oxford, Oxford University Press.
• Povinelli, D. J., & Vonk, J. (2004). “We Don't Need a Microscope to Explore the Chimpanzee Mind.” Mind and Language, 19: 1-28.
• Premack, D. (1971). “On the Assessment of Language Competence in the Chimpanzee.” Behavior of Nonhuman Primates Vol. 4 . A. M. Schrier and F. Stollnitz. New York, Academic Press. 185-228.
• Premack, D., & Woodruff, G. (1978). “Does the Chimpanzee have a Theory of Mind?” Behavioral and Brian Sciences, 1: 515-526.
• Premack, D., & Premack, A. (2003). Original Intelligence: Unlocking the Mystery of Who We Are . New York: McGraw-Hill.
• Premack, D. (2004). “Is Language the Key to Human Intelligence.” Science, 303: 318-320.
• Proust, J. (1999). “Mind, Space and Objectivity in Non-Human Animals.” Erkenntnis, 51: 41-58.
• Proust, J. (2006). “Rationality and Metacognition in Non-Human Animals.” Rational Animals? S. Hurley and M. Nudds. Oxford, Oxford University Press: 247-274.
• Ramsey, G., Bastian, M. L., & van Schaik, C. (2007). “Animal Innovation Defined and Operationalized.” Behavioral and Brain Sciences 30(4): 393-406. .
• Reader, S. M., & Laland, K. N. (2003). “Animal Innovation: An Introduction.” In S. M. Reader & K. M. Laland (Eds.), Animal Innovation . New York: Oxford University Press. 3-35.
• Reiss, D., & Marino, L. (2001). “Mirror Self-Recognition in the Bottlenose Dolphin: A Case of Cognitive Convergence.” Proceedings of the National Academy of Sciences, 98: 5937-5942.
• Rivas, J., & Burghardt, G. (2002). “Crotalomorphism: A Metaphor for Understanding Anthropomorphism by Omission.” In M. Bekoff, C. Allen & G. Burghardt (Eds.), The Cognitive Animal: Empirical and Theoretical Perspectives on Animal Cognition (pp. 9-18). Cambridge, MA: MIT Press.
• Rizzolatti, G., & Craighero, L. (2004). “The Mirror-neuron System.” Annual Review of Neuroscience , 27(169-192).
• Roberts, W. (1997). Principles of Animal Cognition , McGraw-Hill.
• Romanes, G. J. (1970). Animal Intelligence (Reprint of 1882 ed. published by Kegan Paul, Trench & Co., London ed.). Farnborough, England: Gregg International.
• Roth, G., & Dicke, U. (2005). “Evolution of the Brain and Intelligence.” Trends in Cognitive Science , 9(5), 250-257.
• Rumbaugh, D., T. V. Gill, et al. (1973). “Reading and Sentence Completion by a Chimpanzee.” Science, 182: 731-733.
• Russon, A. E., Andrews, K. & Huss B. (2007). “Innovation and the Grain Problem.” Behavioral and Brain Sciences 30(4): 422.
• Sakura, O. (1998). “Similarities and Varieties: A Brief Sketch On The Reception of Darwinism and Sociobiology in Japan.” Biology and Philosophy, 13: 341-357.
• Santos, L. R., Nissen, A. G., et al. (2006). “Rhesus Monkeys, Macaca Mulatta, know what Others Can and Cannot Hear.” Animal Behaviour, 71: 1175-1181.
• Sapolsky, R.M. (2005). “The Influence of Social Hierarchy on Primate Health.” Science , 308: 648-652.
• Savage-Rumbaugh, E. S., Rumbaugh, D., et al. (1980). “Reference: The Linguistic Essential.” Science , 210(4472): 922-925.
• Savage-Rumbaugh, E. S., Rumbaugh, D., et al. (1980). “Do Apes Use Language?” American Scientist, 68: 49-61.
• Savage-Rumbaugh, E. S. (1986). Ape Language: From Conditioned Response to Symbol . New York, Columbia University Press.
• Savage-Rumbaugh, E. S. & Rumbaugh, D. (1998). “Perspectives on Consciousness, Language, and Other Emergent Processes in Apes and Humans.” Toward a Science of Consciousness II . S. Hameroff, A. Kaszniak and A. Scott, MIT Press. 533-549.
• van Schaik, C. P., Ancrenaz, M., Borgen, G., Galdikas, B., Knott, C. D., Singleton, I., Suzuki, A., Utami, S. S., & Merrill, M. (2003). “Orangutan Cultures and the Evolution of Material Culture.” Science, 299: 102-105.
• Schusterman, R. J. & Krieger, K. (1984). “California Sea Lions are Capable of Semantic Comprehension.” The Psychological Record, 34: 3-23.
• Shanker, S., & King, B. (2002). “The Emergence of a New Paradigm in Ape Language Research.” Behavioral and Brain Sciences , 25(5): 605-620.
• Sheets-Johnstone, M. (1992). “Taking Evolution Seriously.” American Philosophical Quarterly, 29: 343-352.
• Shettleworth, S. J. (1998). Cognition, Evolution, and Behavior . New York: Oxford University Press.
• Shettleworth, S. J. & Sutton, J. E. (2006). “Do Animals Know What They Know?” Rational Animals? S. Hurley and M. Nudds. Oxford, Oxford University Press: 235-246.
• Shumaker, R. W. (1997). Observational Learning In Orangutans . George Mason University.
• Shumaker, R. W., & Swartz, K. B. (2002). “When Traditional Methodologies Fail: Cognitive Studies of Great Apes.” In M. Bekoff, C. Allen & G.M. Burdhardt (Eds.), The Cognitive Animal: Empirical and Theoretical Perspectives on Animal Cognition . Cambridge, MA: MIT Press. 335-345.
• Silk, J.B., Brosnan, S.F., Vonk, J., Henrich, J., Povinelli, D.J., Richardson, A.S., Lambeth, S.P., Mascaro, J. & Schapiro, S.J. (2005). “Chimpanzees Are Indifferent To The Welfare Of Unrelated Group Members.” Nature , 437: 1357-1359.
• Skidmore, J. (2001). “Duties to Animals: The Failure of Kant's Moral Theory.” The Journal of Value Inquiry , 35: 541-559.
• Skinner, B. F. (1935). “The Generic Nature of the Concepts of Stimulus and Response.” The Journal of General Psychology , 12: 40-65.
• Slocombe, K. E., & Zuberbühler, K. (2005). “Functionally Referential Communication in a Chimpanzee.” Current Biology , 15: 1779-1784.
• Slocombe, K. E., & Zuberbühler, K. (2006). “Food-Associated Calls in Chimpanzees: Responses to Food Types or Food Preferences?” Animal Behaviour , 72: 989-999.
• Smith, P. (1982). “On Animal Beliefs.” Southern Journal of Philosophy, 20: 503-512.
• Smith, J. D., Schull, J., et al. (1995). “The Uncertain Response in the Bottlenose Dolphin.” Journal of Experimental Psychology General, 124: 391-408.
• Sober, E. (2005). “Comparative Psychology meets Evolutionary Biology: Morgan's Canon and Cladistic Parsimony.” In Lorraine Datson & Gregg Mitman (Eds.), Thinking with Animals: New Perspectives on Anthropomorphism (pp. 85-99). New York: Columbia University Press.
• Sole, L. M., Shettleworth, S. J., & Bennett, P.J. (2003). “Uncertainty in Pigeons.” Psychonomic Bulletin and Review 10: 738-745.
• Spelke, E.S. (1991). “Physical Knowledge in Infancy: Reflections of Piaget's Theory.” In S. Carey & R. Gelman (Eds.), The Epigenesis of Mind: Essays on Biology and Cognition . Hillsdale, NJ: Erlbaum. 37-61.
• Sterelny, K. (2003). Thought in a Hostile World : Blackwell.
• Stich, S. P. (1978). “Do Animals Have Beliefs?” Australasian Journal of Philosophy , 57: 15-28.
• Struhsaker, T. T. (1967). “Auditory Communication Among Vervet Monkeys.” In S. A. Altmann (Ed.), Social Communication among Primates . Chicago: University of Chicago Press. 281-324.
• Swartz, K. B., & Evans, S. (1991). “Not All Chimpanzees Show Self-Recognition.” Primates, 32: 483-496.
• Swartz, K.B., Sarauw, D., & Evans, S. (1999). “Comparative Aspects of Mirror Self-Recognition in Great Apes.” In T. Parker, R. W. Mitchell, and H. L. Miles (Eds.), The Mentalities of Gorillas and Orangutans in Comparative Perspective . 283-294.
• Terrace, H. S., Petitto, L. A., Sanders, R. J., & Bever, T. G. (1979). “Can an Ape Create a Sentence?” Science , 206(4421): 891-902.
• Thorndike, E.L. (1911). Animal Intelligence . New York: Macmillan.
• Tinbergen, N. (1963). “On Aims and Methods of Ethology.” Zeitschrift für Tierpsychologie, 20: 410-433.
• Tomasello, M. (1996). “Do Apes Ape?” In C. Heyes & B. G. Galef (Eds.), Do Apes Ape? . London: Academic. 319-346.
• Tomasello, M. (1998). “Emulation Learning and Cultural Learning.” Behavioral and Brain Sciences, 21: 703-704.
• Tomasello, M. (1999). The Cultural Origins of Human Cognition . Cambridge, MA: Harvard University Press.
• Tomasello, M., & Zuberbühler, K. (2002). “Primate Vocal and Gestural Communication.” In M. Bekoff, C. Allen & G. M. Burghardt (Eds.), The Cognitive Animal: Empirical and Theoretical Perspectives on Animal Cognition . Cambridge, MA: MIT Press.
• Tomasello, M., Call, J., & Hare, B. (2003a). “Chimpanzees Understand Psychological States-The Question is Which Ones and to What Extent.” Trends in Cognitive Science, 7: 153-156.
• Tomasello, M., Call, J., & Hare, B. (2003b). “Chimpanzees Versus Humans: It's Not That Simple.” Trends-in-Cognitive-Sciences, 7: 239-240.
• Tomasello, M., & Carpenter, M. (2005). “The Emergence of Social Cognition in Three Young Chimpanzees.” Monographs of the Society for Research in Child development , 70: 1-131.
• Uller, C. (1997). Origins of Numerical Concepts: A Comparative Study of Human Infants and Nonhuman Primates . Cambridge, MA: MIT Press.
• Uller, C. (2004). “Disposition to Recognize Goals in Infant Chimpanzees.” Animal Cognition, 7: 154-161.
• van Hoof J. (1967). “The Facial Displays of the Catarrhine Monkeys and Apes.” In D. Morris (Ed.), Primate Ethology . Chicago: Aldine. 7-78.
• van Hoof J. (1972). “A Comparative Approach to the Phylogeny of Laughter and Smile.” In R.A. Hinde (Ed.) Non-Verbal Communication . Cambridge University Press. 209-241.
• van Hoof J. (1973). “A Structural Analysis of the Social Behaviour of a Semi-Captive Group of Chimpanzees.” In Cranach, M.V., Vine, I. (Eds.) Social Communication and Movement. European Monographs in Social Psychology 4 . London: Academic. 75-162.
• Voltaire, F. (1984). Philosophical Dictionary (T. Besterman, Trans.). New York: Penguin Classics.
• de Waal, F. (1982). Chimpanzee Politics: Power and Sex Among Apes . London: Jonathan Cape.
• de Waal, F. (1996). Good Natured: The Origins of Right and Wrong in Humans and Other Animals . Cambridge, MA, Harvard University Press.
• de Waal, F. (1999). “Anthropomorphism and Anthropodenial: Consistency in Our Thinking About Human and Other Animals.” Philosophical Topics, 27: 225-280.
• de Waal, F. (2003). “Silent Invasion: Imanishi's Primatology and Cultural Bias in Science.” Animal Cognition, 6: 293-299.
• de Waal, F. (2006). Primates and Philosophers: How Morality Evolved . Princeton: Princeton University Press.
• Warneken, F., Hare, B., Melis, A., Hanus, D., & Tomasello, M. (2007). “Spontaneous Altruism by Chimpanzees and Young Children.” Public Library of Science Biology, 5(7), 1-7.
• Warneken, F. & Tomasello, M. (2006). “Altruistic Helping in Infants and Young Chimpanzees.” Science, 311(5765).
• Whiten, A., Horner, V., Litchfield, C., & Marshall-Pesini, S. (2004). “How do Apes Ape?” Learning and Behaviour, 32: 36-52.
• Wimmer, H., & Perner, J. (1983). “Beliefs About Beliefs: Representation and Constraining Function of Wrong Beliefs in Young Children's Understanding of Deception.” Cognition, 13: 103-128.
• Zuberbühler, K. (2000). “Referential Labeling in Diana Monkeys.” Animal Behavior, 59: 917-927.
• Zuberbühler, K. (2001). “Predator-Specific Alarm Calls in Campbell's Guenons.” Behavioral Ecology and Sociobiology, 50: 414-422.

Other Internet Resources

• The Ape and the Child , videos and text from Kellogg & Kellogg's study
• ChimpFACS , (Facial Action Coding System)
• Chimpanzee Ai , Primate Research Institute, Kyoto University
• Chimp Haven
• The Dolphin Institute
• Dolphins of Monkey Mia Research Foundation
• The Gorilla Foundation
• The Great Ape Trust
• Language Research Center , operated by the Department of Psychology, Georgia State University
• Living Links , at Emory University
• Max Planck Institute Department of Developmental and Comparative Psychology
• Max Planck Institute Department of Primatology
• The Primate Cognition Lab at Columbia University
• PrimateLit Database , bibliographic database for primatology
• Primate Research Institute, Kyoto University
• Project Delphis , dolphin cognition research
• Scottish Primate Research Group
• Smithsonian National Zoological Park Think Tank
• SPAM: Society for the Philosophy of Animal Minds
• University of Cambridge Sub-Department of Animal Behaviour
• Voltaire's philosophical dictionary “ Animals ”
• Wisconsin National Primate Research Center
• Yerkes National Primate Research Center

altruism: biological | animals, moral status of | behaviorism | cognitive science | consciousness: animal | emotion | folk psychology: as a theory | folk psychology: as mental simulation | moral psychology: empirical approaches | other minds | physics: experiment in | practical reason | pragmatics

Animal Behavior/Lek Polygyny

• 1 Lek Polygyny
• 2 Alternative Hypothesis: Kin Selection Hypothesis
• 3 Hotspot Hypothesis
• 4 Hotshot Hypothesis
• 5 Female Preference Hypothesis
• 6 References

Lek Polygyny

Lek polygyny is a mating system common in polygynous species of insects and birds in which the male provides no parental care to its offspring. The lek mating system is uniquely driven by the females’ pursuit of their mate, rather than the males'. Males of lekking species do not hunt for receptive females. Males form aggregates in neutral locations devoid any resources valuable to females. The group of males performs intricate vocal, visual or chemical displays to lure receptive females to their lekking site. In most lekking species, these group displays typically increase the ratio of visiting females per males. At the lekking site, visiting females are able to compare the males’ physiques and courtship displays, picking the most attractive male as their mate (Alcock, 2001). Thus, the few, most attractive males will do the majority of the mating (about 99%), while the subordinate males do no mating at all (Sherman, 1999).

Alternative Hypothesis: Kin Selection Hypothesis

The Lek Polygyny mating system promotes a heavily skewed mating success rate among lekking males. Although most of the individuals in a lek never receive a mating opportunity, lek polygamy continues to flourish among various species of birds and insects (Sherman, 1999). This suggests that the fitness of subordinate males must somehow be indirectly benefited by communal displays. A number of hypotheses have been proposed to explain the reasons behind lekking behavior. Widely recognized hypotheses include the Hotspot Hypothesis, the Hotshot Hypothesis, and the Female Preference Hypothesis (Alcock, 2001). Petrie et al. (1999) proposed an alternative hypothesis predicting that lekking behavior is driven by kin selection.

In the kin selection hypothesis, Petrie et al. (1999) suggests that if all the males in a lek were genetically related, then the males (even the subordinate males) would receive fitness benefits. Petrie et al. (1999) demonstrated the kin selection hypothesis by determining the genetic structure of different lekking groups of the Whipsnade Park peacock population. Peacocks, a typical lekking species, form aggregates at neutral display sites. Peacocks use their calls to attract receptive peahens. Upon the arrival of a peahen, the peacocks cease calling and perform intricate plumage displays. As common in lekking species, the peacocks with the most appealing courtship displays have a high mating success while the subordinate peacocks have no mates. The lekking sites of peacocks are carefully chosen and after a male’s fourth year of lekking, he will establish his permanent lekking site. Peacocks will return to this same lekking site every mating season (Petrie et al. , 1999).

Using multilocus fingerprinting, Petrie et al. (1999) compared the genetic similarity of within and between the lekking groups of Whipsnade Park’s peacocks. The results supported the proposed kin selection hypothesis. As predicted, the degree of band-sharing within the leks was substantially higher than the band-sharing between the leks. The band sharing within the leks was indicative of that of half-siblings. After this discovery, Petrie et al. (1999) proposed subsequent hypotheses of why these lekking groups consisted of related individuals. Petrie et al. (1999) suggested that related peacocks tended to display together because their dispersal was concentrated around their natal sites. This hypothesis was rejected when peacocks of mixed relatedness were reared away from their natal sites but, upon reintroduction into the Whipsnade Park population, joined leks with closely related peacocks. It was then suggested that peacocks’ tendency to congregate with relatives was due to a shared genetic preference for habitat-selection. This was rejected due to the homogenous environmental features of the park. Petrie et al. (1999) concluded that unique structure of the peacock leks was driven by kin selection based on self-referent phenotypic matching (“Armpit Effect” ). Petrie et al. (1994) predicted that peacocks match heritable similarities of their own phenotype with those of other males. Because the authors believed that peacocks do not participate in the rearing of their offspring, they posit that a peafowl’s recognition of their father must be genetically innate. The tendency to form leks with relatives, therefore, occurs in the absence of social cues to their identity. By cooperatively forming leks with genetically related individuals, subordinate peacocks forfeit their chances of mating, but increase the chance that their genes will be passed on via successful relatives. The indirect benefits of fitness seen in peacock lekking displays outweigh the costs of communal displays (Petrie et al. , 1999).

There are four distinct genera of Peafowl: Pavo, Afropavo, Rheiinartia and Argusianus. Though it is frequently reported that male peafowl do not participate in nest defense or rearing paternal care of young has been observed in all four genera. Only the common Indian Peafowl Pavo cristatus is frequently kept in captivity. Its behaviors in zoo settings may not reflect its evolutionary history to the degree that wild populations of the species might.

Hotspot Hypothesis

According to the hotspot hypothesis, males form leks because females frequently visit certain "hotspots".

Hotshot Hypothesis

The hotshot hypothesis predicts that males form leks because subordinate males congregate around highly attractive males to increase their chance of being noticed by receptive females.

Female Preference Hypothesis

The female preference hypothesis predicts that males form leks because female like to visit large clusters of males consisting of a variety of potential mates from which she can quickly and safely compare the quality of her mating choices.

Alcock, J. Animal Behavior . 7th ed. Sunderland: Sinauer Associates, 2002.

Petrie M, Krupa A, Burke T: Peacocks lek with relatives even in the absence of social and environmental cues. Nature 401: 155-157 (1999).

Sherman PW: Bird of a feather flock together. Nature 401:119-120 (1999).

Return to Animal Behavior

• Book:Animal Behavior

Navigation menu

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Author Biographies
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Comparison of the efficacy and welfare of different training methods in stopping chasing behavior in dogs.

Simple Summary

1. introduction, 2. materials and methods, 2.1. participants and location, 2.2. study trainers, 2.3. materials, 2.4. procedure, 2.4.1. day 1, 2.4.2. training, stopping word, group a procedure: day 2–day 4, group b and group c procedure: day 2, word conditioning, group b procedure: day 2, session 2–day 4, group c procedure, day 2, session 2–day 4, 2.4.3. testing, 2.5. cortisol collection, 2.6. video analysis and ethogram, 2.7. data and statistical analysis, 3.1. participants, 3.2. test performance, 3.3. cortisol analysis, 3.4. behavior analysis, 3.5. state behaviors, 4. discussion, 4.1. limitations, 4.2. ethical considerations, 5. conclusions, supplementary materials, author contributions, institutional review board statement, informed consent statement, data availability statement, acknowledgments, conflicts of interest.

• Carter, R.H. Animal Training Device. U.S. Patent No. 2,023,950, 10 December 1935. [ Google Scholar ]
• Cameron, R.C. Radio Controlled Electric Cutaneous Signal Type Animal Obedience Device. U.S. Patent No. 2,800,104, 23 July 1957. [ Google Scholar ]
• Pettingill, R.P. Animal Training Collar. U.S. Patent No. 2,996,043, 15 August 1961. [ Google Scholar ]
• Solomon, R.L.; Wynne, L.C. Traumatic avoidance learning: Acquisition in normal dogs. Psychol. Monogr. Gen. Appl. 1953 , 67 , 1. [ Google Scholar ] [ CrossRef ]
• Christiansen, F.O.; Bakken, M.; Braastad, B.O. Behavioural changes and aversive conditioning in hunting dogs by the second-year confrontation with domestic sheep. Appl. Anim. Behav. Sci. 2001 , 72 , 131–143. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Schilder, M.B.; van der Borg, J.A. Training dogs with help [sic] of the shock collar: Short and long term behavioural effects. Appl. Anim. Behav. Sci. 2004 , 85 , 319–334. [ Google Scholar ] [ CrossRef ]
• Schalke, E.; Stichnoth, J.; Ott, S.; Jones-Baade, R. Clinical signs caused by the use of electric training collars on dogs in everyday life situations. Appl. Anim. Behav. Sci. 2007 , 105 , 369–380. [ Google Scholar ] [ CrossRef ]
• Salgirli, Y.; Schalke, E.; Hackbarth, H. Comparison of learning effects and stress between 3 different training methods (electronic training collar, pinch collar and quitting signal) in Belgian Malinois Police Dogs. Rev. Méd. Véterinaire 2012 , 163 , 530–535. [ Google Scholar ] [ CrossRef ]
• Dale, A.R.; Statham, S.; Podlesnik, C.A.; Elliffe, D. The acquisition and maintenance of dogs’ aversion responses to kiwi ( Apteryx spp.) training stimuli across time and locations. Appl. Anim. Behav. Sci. 2013 , 146 , 107–111. [ Google Scholar ] [ CrossRef ]
• Dale, A.R.; Podlesnik, C.A.; Elliffe, D. Evaluation of an aversion-based program designed to reduce predation of native birds by dogs: An analysis of training records for 1156 dogs. Appl. Anim. Behav. Sci. 2017 , 191 , 59–66. [ Google Scholar ] [ CrossRef ]
• China, L.; Mills, D.S.; Cooper, J.J. Efficacy of Dog Training with and without Remote Electronic Collars vs. a Focus on Positive Reinforcement. Front. Vet. Sci. 2020 , 7 , 508. [ Google Scholar ] [ CrossRef ]
• Cooper, J.J.; Cracknell, N.; Hardiman, J.; Wright, H.; Mills, D. The welfare consequences and efficacy of training pet dogs with remote electronic training collars in comparison to reward based training. PLoS ONE 2014 , 9 , e102722. [ Google Scholar ] [ CrossRef ]
• Sargisson, R.J.; McLean, I.G. Commentary: Efficacy of Dog Training with and without Remote Electronic Collars vs. a Focus on Positive Reinforcement. Front. Vet. Sci. 2021 , 8 , 629746. [ Google Scholar ] [ CrossRef ]
• de Castro, A.C.V.; Barrett, J.; de Sousa, L.; Olsson, I.A.S. Carrots versus sticks: The relationship between training methods and dog-owner attachment. Appl. Anim. Behav. Sci. 2019 , 219 , 104831. [ Google Scholar ] [ CrossRef ]
• Casey, R.A.; Naj-Oleari, M.; Campbell, S.; Mendl, M.; Blackwell, E.J. Dogs are more pessimistic if their owners use two or more aversive training methods. Sci. Rep. 2021 , 11 , 19023. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Vieira de Castro, A.C.; Fuchs, D.; Morello, G.M.; Pastur, S.; de Sousa, L.; Olsson, I.A.S. Does training method matter? Evidence for the negative impact of aversive-based methods on companion dog welfare. PLoS ONE 2020 , 15 , e0225023. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rooney, N.J.; Cowan, S. Training methods and owner-dog interactions: Links with dog behaviour and learning ability. Appl. Anim. Behav. Sci. 2011 , 132 , 169–177. [ Google Scholar ] [ CrossRef ]
• Deldalle, S.; Gaunet, F. Effects of 2 training methods on stress-related behaviors of the dog ( Canis familiaris ) and on the dog–owner relationship. J. Vet. Behav. 2014 , 9 , 58–65. [ Google Scholar ] [ CrossRef ]
• Hiby, E.F.; Rooney, N.J.; Bradshaw, J.W.S. Dog training methods: Their use, effectiveness and interaction with behaviour and welfare. Anim. Welf. 2004 , 13 , 63–70. [ Google Scholar ] [ CrossRef ]
• Blackwell, E.J.; Bolster, C.; Richards, G.; Loftus, B.A.; Casey, R.A. The use of electronic collars for training domestic dogs: Estimated prevalence, reasons and risk factors for use, and owner perceived success as compared to other training methods. BMC Vet. Res. 2012 , 8 , 93. [ Google Scholar ] [ CrossRef ]
• Masson, S.; Nigron, I.; Gaultier, E. Questionnaire survey on the use of different e-collar types in France in everyday life with a view to providing recommendations for possible future regulations. J. Vet. Behav. 2018 , 26 , 48–60. [ Google Scholar ] [ CrossRef ]
• Johnson, T. A Simple Line between Life and Death: The Leash. VetzInsight. 11 October 2012. Available online: https://www.vin.com/vetzinsight/default.aspx?pid=756&id=5563139 (accessed on 29 July 2024).
• Michaelson Found Animals. How to Tame Prey Chase Drive in a Dog. Available online: https://www.foundanimals.org/control-dogs-predatory-chase-drive/ (accessed on 16 March 2024).
• Pappageorge, B. How to Stop Dog Chasing with an E-Collar. Educator Collars. 5 November 2018. Available online: https://www.educatorcollars.com/blog/stop-dog-chasing-e-collar (accessed on 10 March 2024).
• PetSpy. How to Stop a Dog from Chasing with an E-Collar. PetSpy. 16 April 2023. Available online: https://petspy.com/blogs/dog-training/stop-dog-from-chasing-with-ecollar (accessed on 10 March 2024).
• Baril, K.E. Train Your Dog Not to Chase Wildlife. Animal Wellness Magazine. 14 July 2017. Available online: https://animalwellnessmagazine.com/train-dog-not-chase-wildlife (accessed on 10 March 2024).
• Bangura, W. How to Train Your Dog to Stop Chasing: A Comprehensive Guide. 6 February 2024. The Dog Behaviorist. Available online: https://dogbehaviorist.com/2024/02/06/how-to-train-your-dog-to-stop-chasing-a-comprehensive-guide/ (accessed on 10 March 2024).
• Miller, P. Training Your Dog to Execute an “Extremely Fast” Reliable Recall. 15 August 2012. Available online: https://www.whole-dog-journal.com/training/leash_training/training-your-dog-to-execute-an-extremely-fast-reliable-recall/ (accessed on 10 March 2024).
• Kozlowski, C.P.; Clawitter, H.; Guglielmino, A.; Schamel, J.; Baker, S.; Franklin, A.D.; Powell, D.; Coonan, T.J.; Asa, C.S. Factors affecting glucocorticoid and thyroid hormone production of Island foxes. J. Wildl. Manag. 2020 , 84 , 505–514. [ Google Scholar ] [ CrossRef ]
• Friard, O.; Gamba, M. BORIS: A free, versatile open-source event-logging software for video/audio coding and live observations. Methods Ecol. Evol. 2016 , 7 , 1325–1330. [ Google Scholar ] [ CrossRef ]
• Schatz, S.; Palme, R. Measurement of faecal cortisol metabolites in cats and dogs: A non-invasive method for evaluating adrenocortical function. Vet. Res. Commun. 2001 , 25 , 271–287. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Madison, C. How to Teach Your Dog the Emegency Recall. Preventative Vet. 27 May 2021. Available online: https://www.preventivevet.com/dogs/teach-your-dog-the-emergency-recall-come (accessed on 10 March 2024).
• Electronic Collar Manufacturer’s Association (EMCA). FAQ. Available online: http://ecma.eu.com/faq/ (accessed on 23 March 2024).
• Garmin. Garmin Sport Pro: Owner’s Manual. 2015. Available online: https://static.garmin.com/pumac/Sport_PRO_OM_EN.pdf (accessed on 10 March 2024).
• Azrin, N.H.; Holz, W.C. Punishment. In Operant Behavior: Areas of Research and Application ; Honig, W.K., Ed.; Appleton-Century-Crofts: Norwalk, CT, USA, 1966; pp. 380–447. [ Google Scholar ]
• Rankin, C.H.; Abrams, T.; Barry, R.J.; Bhatnagar, S.; Clayton, D.F.; Colombo, J.; Coppola, G.; Geyer, M.A.; Glanzman, D.L.; Marsland, S.; et al. Habituation revisited: An updated and revised description of the behavioral characteristics of habituation. Neurobiol. Learn. Mem. 2009 , 92 , 135–138. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Dinsmoor, J.A. Punishment. In Learning and Behavior Therapy ; O’Donohue, W.T., Ed.; Allyn & Ba: Boston, MA, USA, 1998; pp. 188–204. [ Google Scholar ]
• Hineline, P.N.; Rosales-Ruiz, J. Behavior in relation to aversive events: Punishment and negative reinforcement. In APA Handbook of Behavior Analysis, Vol. 1: Methods and Principles ; Madden, G.J., Dube, W.V., Hackenberg, T.D., Hanley, G.P., Lattal, K.A., Eds.; American Psychiatric Association: Washington, DC, USA, 2013; pp. 483–512. [ Google Scholar ] [ CrossRef ]
• BCSPCA. Emergency Recall—The One World Every Dog Must Know. 25 August 2023. Available online: https://spca.bc.ca/news/emergency-recall/ (accessed on 10 March 2024).
• Marschark, E.D.; Baenninger, R. Modification of instinctive herding dog behavior using reinforcement and punishment. Anthrozoös 2002 , 15 , 51–68. [ Google Scholar ] [ CrossRef ]
• Starling, M.; Spurrett, A.; McGreevy, P. Effects of Lure Type on Chase-related Behavior in Racing Greyhounds. Animals 2020 , 10 , 2262. [ Google Scholar ] [ CrossRef ]
• Accorsi, P.A.; Carloni, E.; Valsecchi, P.; Viggiani, R.; Gamberoni, M.; Tamanini, C.; Seren, E. Cortisol determination in hair and faeces from domestic cats and dogs. Gen. Comp. Endocrinol. 2008 , 155 , 398–402. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Bryan, H.M.; Adams, A.G.; Invik, R.M.; Wynne-Edwards, K.E.; Smits, J.E. Hair as a meaningful measure of baseline cortisol levels over time in dogs. J. Am. Assoc. Lab. Anim. Sci. 2013 , 52 , 189–196. Available online: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3624789 (accessed on 10 March 2024). [ PubMed ]
• Schork, I.G.; Manzo, I.A.; De Oliveira, M.R.B.; Costa, F.V.; Palme, R.; Young, R.J.; De Azevedo, C.S. How environmental conditions affect sleep? An investigation in domestic dogs ( Canis lupus familiaris). Behav. Process. 2022 , 199 , 104662. [ Google Scholar ] [ CrossRef ]
• Slotta-Bachmayr, L.; Schwarzenberger, F. Faecal cortisol metabolites as indicators of stress during training and search missions in avalanche dogs. Vet. Med. Austria 2007 , 64 , 110–117. [ Google Scholar ]
• Uetake, K.; Uchida, H.; Ishihara, J.; Kushima, S.; Tanaka, T. Faecal corticosterone levels of dogs relinquished to a shelter in Japan. Vet. Rec. Open 2015 , 2 , e000139. [ Google Scholar ] [ CrossRef ]
• Zannoni, A.; Pietra, M.; Gaspardo, A.; Accorsi, P.A.; Barone, M.; Turroni, S.; Laghi, L.; Zhu, C.; Brigidi, P.; Forni, M. Non-invasive assessment of fecal stress biomarkers in hunting dogs during exercise and at rest. Front. Vet. Sci. 2020 , 7 , 126. [ Google Scholar ] [ CrossRef ]
• Cobb, M.L.; Iskandarani, K.; Chinchilli, V.M.; Dreschel, N.A. A systematic review and meta-analysis of salivary cortisol measurement in domestic canines. Domest. Anim. Endocrinol. 2016 , 57 , 31–42. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Beerda, B.; Schilder, M.B.; van Hooff, J.A.; de Vries, H.W. Manifestations of chronic and acute stress in dogs. Appl. Anim. Behav. Sci. 1997 , 52 , 307–319. [ Google Scholar ] [ CrossRef ]
• Pappageorge, B. 5 Reasons to Use an E-Collar to Train Your Dog. Educator Collars. 15 September 2017. Available online: https://www.educatorcollars.com/blog/remote-training-collar#:~:text=A%20remote%20training%20collar%20gives%20you%20a%20way%20to%20interrupt,that%20they%20can%20learn%20faster (accessed on 10 March 2024).
• Siler, W. How a Shock Collar Changed My Relationship with My Dog. Outside Magazine. 6 October 2022. Available online: https://www.outsideonline.com/outdoor-gear/tools/dog-shock-collar-training-experience/ (accessed on 10 March 2024).
• Domjan, M. Pavlovian conditioning: A functional perspective. Annu. Rev. Psychol. 2005 , 56 , 179–206. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Feuerbacher, E.N.; Wynne, C.D. Most domestic dogs ( Canis lupus familiaris) prefer food to petting: Population, context, and schedule effects in concurrent choice. J. Exp. Anal. Behav. 2014 , 101 , 385–405. [ Google Scholar ] [ CrossRef ]
• McLennan, T. Does the attention a dog pays to their owner increase after the dog engages in activities that mimic the predatory preferences of that dog ( Canis familiaris )? Appl. Anim. Behav. Sci. 2023 , 263 , 105944. [ Google Scholar ] [ CrossRef ]
• Herron, M.E.; Shofer, F.S.; Reisner, I.R. Survey of the use and outcome of confrontational and non-confrontational training methods in client-owned dogs showing undesired behaviors. Appl. Anim. Behav. Sci. 2009 , 117 , 47–54. [ Google Scholar ] [ CrossRef ]
• Makowska, I.J.; Cavalli, C.M. Review of Dog Training Methods: Welfare, Learning Ability, and Current Standards ; British Columbia Society for the Prevention of Cruelty of Animals: Vancouver, BC, Canada, 2023. [ Google Scholar ]
• Arkow, P.S.; Dow, S. Ties that do not bind: A study of the human-animal bonds that fail. In Pet Connection: Its Influence on Our Health and Quality of Life ; Anderson, R.K., Allen, B.L., Hart, L.A., Eds.; Center to Study Human-Animal Relationships and Environments: Columbia, MO, USA, 1984. [ Google Scholar ]
• Wells, D.L.; Hepper, P.G. Prevalence of behaviour problems reported by owners of dogs purchased from an animal rescue shelter. Appl. Anim. Behav. Sci. 2000 , 69 , 55–65. [ Google Scholar ] [ CrossRef ]
• Lindsay, S.R. (Ed.) Handbook of Applied Dog Behavior and Training, Procedures and Protocols ; John Wiley & Sons: Hoboken, NJ, USA, 2013; Volume 3. [ Google Scholar ]
• Fernandez, E.J. The least inhibitive, functionally effective model: A new framework for ethical animal training practices. J. Vet. Behav. 2024 , 71 , 63–68. [ Google Scholar ] [ CrossRef ]
• Johnson, A.C.; Wynne, C.D. Training Dogs with Science or with Nature? An Exploration of Trainers’ Word Use, Gender, and Certification Across Dog-Training Methods. Anthrozoös 2022 , 36 , 35–51. [ Google Scholar ] [ CrossRef ]
• Blackwell, E.J.; Twells, C.; Seawright, A.; Casey, R.A. The relationship between training methods and the occurrence of behavior problems, as reported by owners, in a population of domestic dogs. J. Vet. Behav. 2008 , 3 , 207–217. [ Google Scholar ] [ CrossRef ]
• Johnson, A.C.; Wynne, C.D. Training Methods Used by Dog Guardians in the United States: Prevalence, Sources of Information, and Reasons for Use. Animals 2024 , 14 , 1310. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Verified Market Reports. E-Collar for Dog Market Size, Share and Growth. Verified Market Reports. December 2023. Available online: https://www.verifiedmarketreports.com/product/e-collar-for-dog-market/ (accessed on 29 July 2024).
• Hernandez, V. Radio Systems Corporation. EDCV 22-1861 JGB. 28 US #312. 2023. Available online: https://www.classaction.org/media/hernandez-v-radio-systems-corporation.pdf (accessed on 10 March 2024).

Click here to enlarge figure

Share and Cite

Johnson, A.C.; Wynne, C.D.L. Comparison of the Efficacy and Welfare of Different Training Methods in Stopping Chasing Behavior in Dogs. Animals 2024 , 14 , 2632. https://doi.org/10.3390/ani14182632

Johnson AC, Wynne CDL. Comparison of the Efficacy and Welfare of Different Training Methods in Stopping Chasing Behavior in Dogs. Animals . 2024; 14(18):2632. https://doi.org/10.3390/ani14182632

Johnson, Anamarie C., and Clive D. L. Wynne. 2024. "Comparison of the Efficacy and Welfare of Different Training Methods in Stopping Chasing Behavior in Dogs" Animals 14, no. 18: 2632. https://doi.org/10.3390/ani14182632

Article Metrics

Article access statistics, supplementary material.

ZIP-Document (ZIP, 150 KiB)

Further Information

Mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals