> For the complete documentation index, see [llms.txt](https://manymanys.gitbook.io/mm1-lab-manual/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://manymanys.gitbook.io/mm1-lab-manual/study-background/rationale-and-aims.md).

# Rationale and aims

## Introduction

Understanding variation in cognitive abilities across animal lineages requires integrating both proximate mechanisms and ultimate causes ([Mayr, 1961](https://www.science.org/doi/abs/10.1126/science.134.3489.1501); [Tinbergen, 1963](https://onlinelibrary.wiley.com/doi/10.1111/j.1439-0310.1963.tb01161.x)). These two levels of explanation address different but complementary questions about why a trait exists and how it functions. Proximate mechanisms answer “how does it work?” questions, by studying the immediate biological processes that produce a behavior or cognitive ability. At this level, differences in cognitive performance have been linked both to between-species variation in neural architecture—such as brain size and cortical organization—and to within-species variation in ontogeny and experience ([Boogert et al., 2018](https://royalsocietypublishing.org/rstb/article/373/1756/20170280/30419); [Thornton & Lukas, 2012](https://pmc.ncbi.nlm.nih.gov/articles/PMC3427550/)). By contrast, ultimate causes address “why does it exist?” questions by focusing on the adaptive value and evolutionary history of a trait. At this level, selective social and environmental pressures have been proposed as key drivers of cognitive evolution ([Reader & Laland, 2002](https://www.pnas.org/doi/pdf/10.1073/pnas.062041299)). These include challenges arising from social living ([Dunbar, 1998](https://www.cognitionandculture.net/wp-content/uploads/Evolutionary-Anthropology-1998-Dunbar-The-social-brain-hypothesis.pdf); [Humphrey, 1976](https://web-archive.southampton.ac.uk/cogprints.org/2694/1/SocialFunctionTxt.pdf?ncid=txtlnkusaolp00000618); [Jolly, 1966](https://www.science.org/doi/abs/10.1126/science.153.3735.501)), foraging demands ([Barton, 2012](https://royalsocietypublishing.org/rstb/article-abstract/367/1599/2097/21927/Embodied-cognitive-evolution-and-the?redirectedFrom=fulltext); [Byrne, 1997](https://psycnet.apa.org/record/1997-36638-010); [Parker, 2015](https://www.sciencedirect.com/science/article/abs/pii/S0732118X1400049X); [Parker & Gibson, 1977](https://www.sciencedirect.com/science/article/abs/pii/S0047248477801358)), and environmental unpredictability ([Allman et al., 1993](https://www.pnas.org/doi/pdf/10.1073/pnas.90.1.118); [Sol, 2009](https://pubmed.ncbi.nlm.nih.gov/19049952/)).

Many of the hypotheses about cognition introduced above were developed and tested within a relatively narrow taxonomic range—most prominently large-brained vertebrates such as primates. This restricted focus risks conflating lineage-specific characteristics with general evolutionary principles and overlooking non-mainstream neural architectures and cognitive strategies, thereby introducing an anthropocentric bias ([Shettleworth, 1998/2010](https://psycnet.apa.org/record/2009-24069-000)).

The comparative approach addresses this bias by treating cognition as a set of traits that vary systematically across species and ecological contexts, allowing researchers to examine how underlying mechanisms interact with environmental demands and evolutionary history to produce both shared and divergent outcomes ([Beran et al., 2014](https://pmc.ncbi.nlm.nih.gov/articles/PMC4239033/); [Bitterman, 1960](https://psycnet.apa.org/record/1962-02624-001); [Mackintosh, 1988](https://bpspsychub.onlinelibrary.wiley.com/doi/abs/10.1111/j.2044-8295.1988.tb02749.x); [Macphail, 1987](https://psycnet.apa.org/record/1989-10913-001); [Shettleworth, 2012](https://global.oup.com/academic/product/fundamentals-of-comparative-cognition-9780195343106?cc=us\&lang=en&); [Watson, 1914](https://psycnet.apa.org/record/2005-12950-000); [Zentall, 2023](https://www.mdpi.com/2076-2615/13/7/1165)). By expanding analyses across a broader diversity of taxa, comparative research enables the identification of conserved neural processes, reveals convergent evolutionary solutions to similar adaptive challenges, and provides a critical baseline for situating human cognition within a continuous evolutionary landscape rather than as an isolated endpoint ([Shettleworth, 1998/2010](https://psycnet.apa.org/record/2009-24069-000)). Within this framework, researchers often focus on specific cognitive traits that can be measured across species and linked to both underlying mechanisms and ecological demands. One such trait that has received considerable attention is behavioral flexibility.

*Behavioral flexibility*—the ability to modify actions in response to shifting environmental conditions ([Jones, 2005](https://link.springer.com/chapter/10.1007/978-0-387-23327-7_9))—is a core component of cognitive adaptability and executive functions ([Diamond, 2013](https://www.annualreviews.org/content/journals/10.1146/annurev-psych-113011-143750)). It underlies key functions such as exploiting novel resources, evading threats, generating solutions to novel problems, and adjusting to dynamic social environments, capacities that are evolutionarily advantageous and contribute directly to survival and reproductive success (see [Mazza & Šlipogor, 2024](https://academic.oup.com/cz/article/70/3/304/7686939?guestAccessKey=)). A standard experimental procedure used to quantify behavioral flexibility across species is reversal learning ([Harlow, 1949](https://psycnet.apa.org/record/1949-03097-001); [Izquierdo et al., 2017](https://www.sciencedirect.com/science/article/abs/pii/S030645221600244X); [Rajalakshmi & Jeeves, 1965](https://www.sciencedirect.com/science/article/abs/pii/0003347265900357)). In its basic form, subjects are first trained to discriminate between two stimuli, one of which is associated with a reward (S+), while the other is not (S−). Once a predefined learning criterion for selecting S+ is met, the contingencies are reversed: the previously rewarded stimulus becomes unrewarded (S−), and the previously unrewarded stimulus becomes the newly rewarded stimulus (S+). Successful performance thus requires both response inhibition and the acquisition of a novel stimulus-reward mapping. Iterated or interreversal versions of the task, in which multiple reversals are introduced, allow researchers to examine the extent to which subjects acquire learning sets—that is, whether they improve across successive reversals.

Despite decades of research on reversal learning, the field currently lacks a common experimental framework. Most studies rely on species-specific tasks and distinct stimuli, procedures, and performance measures, making it nearly impossible to draw robust species comparisons or uncover general principles of behavioral flexibility.

## Proximate-level explanations of reversal learning

### Theory 1: Associative Learning and Reinforcement

From both classical conditioning and operant conditioning perspectives, reversal learning involves updating previously learned associations that are either stimulus-driven ([Rescorla & Wagner, 1972](https://www.researchgate.net/publication/239030972_A_theory_of_Pavlovian_conditioning_The_effectiveness_of_reinforcement_and_non-reinforcement)) or behavior-outcome driven (Skinner, [1938](https://www.bfskinner.org/wp-content/uploads/2016/02/BoO.pdf), [1953](https://www.bfskinner.org/newtestsite/wp-content/uploads/2014/02/ScienceHumanBehavior.pdf)). Because these mechanisms are considered general associative learning processes, both classical and operant conditioning frameworks predict similar patterns, with reversal taking longer than initial acquisition, individuals showing perseverative errors immediately after contingency switch, and performance gradually adjusting based on reinforcement feedback. Different patterns may also arise from differences in learning rates, sensitivity to reinforcement and nonreinforcement, stimulus salience, perceptual abilities, and motivational factors.

### Theory 2: Inhibitory Control and Executive Function

An alternative but complementary account emphasizes inhibitory control mechanisms ([Diamond, 2013](https://www.annualreviews.org/content/journals/10.1146/annurev-psych-113011-143750); [Shettleworth, 2010](https://psycnet.apa.org/record/2009-24069-000)). According to this view, reversal errors should cluster immediately after contingency change, reflecting perseveration of the previously rewarded response. Subjects who acquire rapidly but fail in reversal would show a dissociation consistent with executive-control limitations.

### Theory 3: Attention

Attentional models propose that organisms do not treat all stimuli equally during learning ([Pearce & Mackintosh, 2010](https://www.google.com/books/edition/Attention_and_Associative_Learning/wn-hHPt_ftgC?hl=en\&gbpv=1\&dq=Pierce+and+Hall+attentional+theory\&pg=PA11\&printsec=frontcover)). Instead, they allocate attention selectively to cues that are either good predictors of outcomes (learned predictiveness; [Mackintosh, 1975](https://psycnet.apa.org/record/1975-26802-001)) or, by contrast, when the outcomes are surprising or uncertain ([Pearce & Hall, 1980](https://psycnet.apa.org/record/1981-02676-001)). Mackintosh further argued that these attentional processes are ecologically adaptive, allowing animals to prioritize cues that reliably signal biologically relevant events in complex and variable environments ([Mackintosh, 1983](https://cir.nii.ac.jp/crid/1970304959848305984)). Mackintosh’s original model predicts early perseveration after reversal due to attention to previously predictive cues (i.e., subjects should learn the initial phase more quickly), whereas the Pearce–Hall model predicts that surprise at reversal increases attention and facilitates reversal learning (i.e., subjects should learn the reversal phase more quickly).

### Theory 4: Frustration and Affective Response

Frustration theory proposes that the omission of an expected reward induces an aversive motivational state that influences subsequent behavior ([Amsel, 1992](https://psycnet.apa.org/record/1992-98682-000)). According to this view, immediately after reversal, subjects may show transient increases in response disruption (e.g., NA trials or latency increases). Subjects with higher impulsivity or prepotent response are expected to exhibit greater disruption (e.g., increased NA responses) immediately after reversal and switch faster if frustration accelerates abandonment of the old rule. Alternatively, highly stress-sensitive individuals may disengage, slowing reversal.

## Ultimate-level explanations of reversal learning

### Theory 5: Social Brain

The Social Brain Hypothesis proposes that selection for managing complex social relationships (e.g., coalitions, dominance hierarchies, deception, cooperation) drove the evolution of larger brains and advanced cognitive capacities ([Dunbar, 1998](https://www.cognitionandculture.net/wp-content/uploads/Evolutionary-Anthropology-1998-Dunbar-The-social-brain-hypothesis.pdf); [Humphrey, 1976](https://web-archive.southampton.ac.uk/cogprints.org/2694/1/SocialFunctionTxt.pdf?ncid=txtlnkusaolp00000618); [Jolly, 1966](https://www.science.org/doi/abs/10.1126/science.153.3735.501)). Behavioral flexibility, including reversal learning, is expected to show in situations where individuals must rapidly update behavior in response to changing social contingencies. Reversal performance is expected to reflect social complexity and brain size: highly social, large-brained animals are predicted to show greater behavioral flexibility, whereas less social, smaller-brained animals are likely to exhibit lower reversal efficiency.

### Theory 6: Ecological Intelligence

Unlike the Social Brain Hypothesis, which emphasizes social complexity as the primary driver of cognitive evolution, the Ecological Intelligence Hypothesis emphasizes ecological and foraging challenges as the main selective pressures shaping cognitive abilities ([Clutton-Brock & Harvey, 1980](https://zslpublications.onlinelibrary.wiley.com/doi/abs/10.1111/j.1469-7998.1980.tb01430.x); [Milton, 1981](https://anthrosource.onlinelibrary.wiley.com/doi/abs/10.1525/aa.1981.83.3.02a00020)). Under this broad framework, selection favors enhanced capacities to navigate ecological demands such as extractive foraging, seasonal resource variability, and spatial navigation in heterogeneous habitats. Specific examples of ecological intelligence include the Extractive Foraging (or Technical Intelligence) Hypothesis—which argues that cognitive complexity evolved to solve ecological problems involving embedded or mechanically challenging food resources, tool use, and object manipulation ([Barton, 2012](https://royalsocietypublishing.org/rstb/article-abstract/367/1599/2097/21927/Embodied-cognitive-evolution-and-the?redirectedFrom=fulltext); [Byrne, 1997](https://psycnet.apa.org/record/1997-36638-010); [Parker, 2015](https://www.sciencedirect.com/science/article/abs/pii/S0732118X1400049X); [Parker & Gibson, 1977](https://www.sciencedirect.com/science/article/abs/pii/S0047248477801358))—and the Cognitive Buffer Hypothesis—which posits that enlarged brains evolved to buffer organisms against environmental challenges, thus increasing survival rates and favoring a longer reproductive life ([Allman et al., 1993](https://www.pnas.org/doi/pdf/10.1073/pnas.90.1.118); [Sol, 2009](https://pubmed.ncbi.nlm.nih.gov/19049952/)). Reversal performance is expected to reflect ecological complexity, with animals from richer, more variable, or unpredictable environments typically showing greater behavioral flexibility, whereas those from poorer, simpler, or more stable habitats are predicted to be less flexible.

## Aims

ManyManys 1 (MM1) seeks to develop, validate, and deploy a standardized reversal learning task suitable for broad cross-species comparison. 

Although reversal learning has been studied extensively, prior research has typically focused on single-species paradigms with divergent methodologies, limiting their direct comparability. MM1 addresses this gap by implementing a unified protocol across a wide taxonomic range, enabling systematic assessment of interspecies variation in behavioral flexibility under controlled and comparable conditions. 

This comparative framework will allow us to assess whether reversal learning performance differs across taxa, and to identify potential correlates underlying such differences. MM1 will generate a large dataset that can serve as a foundation for testing hypotheses about the evolution of behavioral flexibility (for a similar approach, see [MacLean et al., 2014](https://www.pnas.org/doi/10.1073/pnas.1323533111)).


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://manymanys.gitbook.io/mm1-lab-manual/study-background/rationale-and-aims.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.