Linköping University | Department of Physics, Chemistry and Biology
Bachelor’s thesis, 16 ECTS | Biology
2020 | LIU-IFM/LITH-EX-G--20/3858--SE
Cognitive Judgement Bias as an
Indicator for Animal Welfare
Emma Ackelman
Supervisor: Per Jensen Examiner: Matthias Laska
Datum
Date
16-06-20
Avdelning, institution Division, Department
Department of Physics, Chemistry and Biology
Linköping University
URL för elektronisk version
ISBN
ISRN: LITH-IFM-G-EX--20/3858--SE
_________________________________________________________________
Serietitel och serienummer ISSN
Title of series, numbering ______________________________
Språk Language Svenska/Swedish Engelska/English ________________ Rapporttyp Report category Licentiatavhandling Examensarbete C-uppsats D-uppsats Övrig rapport _____________ Titel Title
Cognitivie Judgement Bias as an Indicator for Animal Welfare
Författare
Author Emma Ackelman
Nyckelord
Keyword
“Animal welfare”, “Cognitive bias”, “Cognitive judgment bias”, “Emotions in animals” and “Farm animals”.
Sammanfattning
Abstract
Animal welfare has long been a subject under debate. Since animals are unable to voice concerns about their living standards it is of interest to find other ways to secure their wellbeing. A new measurement has been introduced where animals’ own judgement can act as an indicator for their emotional state and welfare. This study summarizes key elements from previous literature and research in order to explain the connection between welfare and judgement bias. Emotions have been defined as either an observable reaction to a stimulus or a subjective conscious experience of the stimulus. The second has been difficult to assess in animals since they cannot vocalise their own interpretations, hence emotions in animals are assessed based on the first definition. The study of animal welfare is in short the study of animal’s judgement of the world, which in turn indicate how animals feel. Cognitive judgement bias has been defined as whether an animal assess an ambiguous stimulus as negative or positive, a common method used to demonstrate this concept is the go no-go method. Animals learn to discriminate between two stimuli and is then presented with an ambiguous stimulus. The response to the ambiguous stimulus is recorded and determine if the animal is optimistic or pessimistic in its judgement. Research has been rather successful in determining factors which can affect animal welfare, opening up for deeper discussions concerning animal cognition, awareness and their effect on welfare, but further refinements are required to assess the influence of judgement bias.
Upphovsrätt
Detta dokument hålls tillgängligt på Internet - eller dess framtida ersättare - under 25 år från publiceringsdatum under förutsättning att inga extraordinära omständigheter uppstår. Tillgång till dokumentet innebär tillstånd för var och en att läsa, ladda ner, skriva ut enstaka kopior för enskilt bruk och att använda det oförändrat för ickekommersiell forskning och för undervisning. Överföring av upphovsrätten vid en senare tidpunkt kan inte upphäva detta tillstånd. All annan användning av dokumentet kräver upphovsmannens medgivande. För att garantera äktheten, säk-erheten och tillgängligheten finns lösningar av teknisk och administrativ art. Upphovsmannens ideella rätt innefattar rätt att bli nämnd som upphovsman i den omfattning som god sed kräver vid användning av dokumentet på ovan beskrivna sätt samt skydd mot att dokumentet ändras eller presenteras i sådan form eller i sådant sammanhang som är kränkande för upphovsmannens litterära eller konstnärliga anseende eller egenart. För ytterligare information om Linköping Uni-versity Electronic Press se förlagets hemsida http://www.ep.liu.se/.
Copyright
The publishers will keep this document online on the Internet - or its possible replacement - for a period of 25 years starting from the date of publication barring exceptional circumstances. The on-line availability of the document implies permanent permission for anyone to read, to download, or to print out single copies for his/hers own use and to use it unchanged for non-commercial research and educational purpose. Subsequent transfers of copyright cannot revoke this permission. All other uses of the document are conditional upon the consent of the copyright owner. The publisher has taken technical and administrative measures to assure authenticity, security and accessibility. According to intellectual property law the author has the right to be mentioned when his/her work is accessed as described above and to be protected against infringement. For additional informa-tion about the Linköping University Electronic Press and its procedures for publicainforma-tion and for
Contents
1 Abstract 1
2 Introduction 2
2.1 Aim . . . 3
3 Methods and Materials 4
3.1 Limitations . . . 4
4 Emotions 5
5 Emotions in animals 6
6 Welfare 8
7 Cognitive bias 9
8 Cognitive judgement bias 11
8.1 Judgement bias test in laboratory animals . . . 11 8.2 Judgement bias test in farm animals . . . 13 8.3 Judgement bias in other animal groups . . . 14
9 Discussion 16
9.1 Ethical considerations . . . 18 9.2 Social considerations . . . 18
1 Abstract
Animal welfare has long been a subject under debate. Since animals are unable to voice concerns about their living standards it is of interest to find other ways to secure their well-being. A new measurement has been introduced where animals’ own judgement can act as an indicator for their emotional state and welfare. This study summarises key elements from previous literature and research in order to explain the connection between welfare and judgement bias. Emotions have been defined as either an observable reaction to a stimulus or a subjective conscious experience of the stimulus. The second has been difficult to asses in animals since they cannot vocalise their own interpretations, hence emotions in animals are assessed based on the first definition. The study of animal welfare is in short the study of animal’s judgement of the world, which in turn indicate how animals feel. Cognitive judgement bias has been defined as whether an animal assess an ambiguous stimulus as negative or positive, a common method used to demonstrate this concept is the go no-go method. Animals learn to discriminate between two stimuli and is then presented with an ambiguous stimulus. The response to the ambiguous stimulus is recorded and determine if the animal is optimistic or pessimistic in its judgement. Research has been rather successful in determining factors which can affect animal welfare, opening up for deeper discussions concerning animal cognition, awareness and their effect on welfare, but further refinements are required to assess the influence of judgement bias.
2 Introduction
From the very beginning when people first started keeping animals in captivity, the focus has mainly been on how the animals should be kept alive, contained and healthy. Good health and good welfare have thus long been considered equivalents. Only in recent years has the influence of behaviour come in the light of welfare (Keeling and Jensen, 2009). Animal welfare is an important aspect in animal housing and can regulate how animals are treated, with more knowledge about the nature of animals we can construct facilities and housing which optimise the living conditions for each species.
Animal welfare can be defined in two different ways. First, by focusing on the biological aspect of the relationship between an individual and its environment being in balance with each other and thus in harmony (Désiré, Boissy, and Veissier, 2002). The second way is to link welfare to what animals feel and thereby how emotions affect an individual’s chances of survival (Keeling and Jensen, 2009). Is our current welfare system optimised for the well-being of all animals we are housing or are there still aspects we have not yet considered?
Emotions, a concept usually in connection with reaction. Emotions in humans can be expressed through behavioural, verbal and psychological responses. On the other hand, animal expression through verbal affirmation is not an option and emotions can only be perceived through animal behaviour and psychological responses to a situation (Désiré et al., 2002). If it is possible to asses the emotions of an animal it can grant further knowledge about animal welfare and how emotional states can be implemented to improve living standards for animals (Désiré et al., 2002). Perhaps other animal species have more in common with humans than we once thought.
Which methods has been used so far to determine animal emotions? Within psychology the term “cognitive bias” accompanies the perception that emotions can effect our view of the world. Cognitive bias is when an individual’s subjective image of its outside world differs from reality which can be affected by their emotional state. Cognitive bias can therefore be explained as how each individual interpret a stimulus in different ways based on their current emotional state (Haselton, Nettle, and Andrews, 2005; Mendl, Burman, Parker, and Paul, 2009).
The same applies to animals where an animal’s assessment of a situation or another animal differs from what is expected. This discrepancy may be due to factors such as emotional responses to a stimulus (Haselton et al., 2005).
One of the most classic tests for animal emotion states is a cognitive judgement bias tests. A cognitive judgement bias is when an individual makes a decision based on their interpretation of
a ambiguous stimulus (Bateson and Nettle, 2015) as well as which expected consequences their actions might have. The later was shown by Boleij and colleagues (2012) when assessing judgement bias in mice. When an individual shall assess a certain situation they rarely have access to models to predict the probabilities of different outcomes, instead they rely on intuitive judgement (estimation or prediction) (Haselton et al., 2005). In short, a cognitive judgement bias test can act as an indicator for aberrations in animal emotions.
A classic example of how judgemental-cognitive bias is measured involves a test subject which learns that two different stimuli result in different outcomes, often one with reward and one without. An intermediate stimulus or signal is then presented to the test subject who then perform a measurable assessment of the situation (Clegg, 2018). The assessment can for example be to approach an object (Douglas, Bateson, Walsh, Bédué, and Edwards, 2012). The intermediate stimulus occur either without or with an occasional reward (Clegg, 2018). The measurable assessment is either seen as an optimistic or pessimistic evaluation of the stimulus (Doyle, Fisher, Hinch, Boissy, and Lee, 2010). Optimism in animals can be defined as an expectation of a higher probability of a favourable outcome, either as a reward or a decreased probability of punishment, whereas pessimism is defined as an expectation for the outcome to be more unfavourable, higher risk for punishment and less chance of reward (Bateson, 2016).
The focus of this study will be on judgement-cognitive bias, and more specifically, how animal welfare is affected by their decisions and behaviour in relation to different stimuli. First the con-nection between emotions in animals and humans shall be assessed since most studies on emotions were first performed on humans. Secondly the connection between emotion and animal welfare shall be assessed, afterwards how judgement bias can act as an indicator for emotional states will be brought up and lastly we shall discuss if cognitive judgement bias test can be utilised to indicate animal welfare.
2.1 Aim
The deviations in animals’ instinctive or learned responses to various stimuli, after being in a favourable or disadvantageous environment, should be able to act as an indication of welfare since behaviour and decisions of animals are affected by their emotional state. The aim of this study is to assemble results from previous studies and reflections in order to get an understanding of how cognitive judgement bias can give us an insight into the welfare of animals and thus how they feel, by assessing how an animal’s judgement is influenced by their emotional state.
3 Methods and Materials
The search engines used mostly include literature found through Linköping University’s search engine UniSearch where the results can be limited by year, whether the text has been peer-reviewed or not, whether it is books or articles requested as well as limit the search to certain subject areas or keywords. Google Scholar was used when a relevant source could not be found through Unisearch. Sources such as the National Encyclopedia and Wikipedia are used in such a way that they name a number of first-hand sources that are then borrowed from libraries or searched for through the university’s search engine. All sources that indicate another source are reviewed and first hand sources are acquired if possible. Sources published from the year 2000 and forward were prioritised if not a more relevant source or a first hand source was found instead.
The keywords used include; “Animal welfare”, “Cognitive bias”, “Cognitive judgement bias”, “Emotions in animals” and “farm animals”. Many words were used in combinations to limit the search to relevant studies, otherwise there would be results ranging from all species and methods which would only lead to confusion. Studies with a focus on livestock species or laboratory animals were primarily researched.
3.1 Limitations
The study is limited to agricultural animals and laboratory animals where most previous studies have been carried out. Links to studies on other animal groups may occur but mainly as reference material. The study is also limited to judgement-cognitive bias, which means how the animals’ behaviour affect their judgement and decisions in different situations, as well as what can influence the behaviour.
4 Emotions
What is the definition of an emotion? Emotions can be explained in a psychological and behavioural sense, or as a subjective conscious experience (Dawkins, 2000). The first refers to observable re-sponses to stimuli, such as the presence or absence of food. The second refers to how we subjectively interpret a given situation (Dawkins, 2000).
There are several different ways to describe emotions. De Waal (2011) describes emotions as “/.../ a temporary state brought about by biologically relevant external stimuli, whether aversive or attractive.” And also explained how they manifest, “ The emotion is marked by specific changes in the organism’s body and mind /.../ Which emotion is triggered is often predictable by the situation in which the organism finds itself /.../ Emotions combine with individual experience and cognitive assessment of the situation to prepare the organism for an optimal response.”
Within humans three systems have been identified to “underlie” emotions. 1. Verbal/Cognitive
2. Autonomic
3. Expressive/Behaviour
The first being verbal and/or cognitive which is often displayed by people telling each other how they feel. The second consists of autonomic responses such as an increased heart rate or rise in hormone levels due to the effect of emotions. The last and third one consists of facial expressions and behaviour changes resulting from emotions (Oatley, Keltner, and Jenkins, 2006).
5 Emotions in animals
When it comes to non-human animals, there has been a long debate on whether animals can feel emotions or not as well as which animals it applies to (Rollin, 1989).
To test for emotions in animals we can not utilise the same mechanisms and tests performed on humans. In face of the three underlying systems of emotions, the first can not be applied to non-human species since they are unable to vocalise their feelings and emotions. The second and third system however can act as template for similarities (Dawkins, 2000). Though it should be kept in mind that these systems are not optimal since several different emotions can result in the same physiological responses, for example the action of running occur both in dangerous and exciting situations (Dawkins, 2000).
Psychologist Bentham (1823) came up with the well known lines "The question is not ’Can they reason?’ nor ’Can they talk?’ but ’Can they suffer?’". To answer this question scientists have earlier tested several aspects of suffering, one being physical pain. Many tests concerning physical pain has used withdrawal reflexes, vocalisation or escape behaviour to indicate pain in animals (Xie, 2011). Pain has been defined as “/.../ an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage”, by Merskey (1979), indicating that emotional pain is another important aspect of suffering. The debate on how emotional pain should be assessed and weighed against physical pain is still ongoing (Rollin, 1989).
Haga and Ranheim (2005) conducted an experiment where they tested the effect of the anaesthetic lidocaine on the sensory nervous system’s response to harm, from the action of castration of piglets. A group of piglets received lidocaine before castration, while a control group did not receive any anaesthetic. Haga and Ranheim (2005) concluded that the animals suffered less when injected with lidocaine since the measured blood pressure, pulse and electroencephalography differed severely between the experimental group and the control group.
Can animals sense emotions in each other? A behavioural study performed by Miller and colleagues (1963) tested if rhesus macaques (Macaca mulatta) can prevent (or neutralise) an unpleasant stimulus, when exposed to an emotional expression change in another macaque. The macaques were first taught how to avoid the unpleasant stimulus themselves by pushing a lever. They were later paired two and two where one of the macaques had access to the lever but it was the other macaque who experienced the unpleasant stimulus in the form of a small chock. The macaque with the lever had to determine how their partner was feeling based on their expressions. Miller and colleagues (1963) concluded that rhesus macaques (Macaca mulatta) had the ability to
indicate different emotional expressions when observing the facial expressions of other individuals by pushing the lever when their companion expressed pain.
6 Welfare
In both Haga and Ranheim (2005) test as well as Miller and colleagues (1963)’s experiment there is little consideration taken to how the animals might suffer from the tests themselves. The focus of this study is to evaluate new methods focusing to use cognition as a means for indicating animal welfare. To be able to understand how cognition can affect an animal’s well-being we first have to understand how there is a connection between cognition, emotions and animal welfare.
The study of emotions can be seen as an aspect of the study of animal welfare since welfare is a representation of how the animal experiences its surroundings (Beausoleil, Stratton, Guesgen, Sutherland, and Johnson, 2016).
If animals can feel emotions and to what degree has long been debated (Rollin, 1989). Désiré and colleagues (2002) concluded that there is evidence indicating criteria for emotions in farm animals, but that further research of these criteria should be performed to assess the variety of emotions the animals can feel.
The question then turns toward how emotions affect the welfare of the animals. The five freedoms of animal welfare was defined by the UK Animal Welfare Council (AWC) in 1979, when they were still named Farm Animal Welfare Council (FAWC). Their report was a result from an investigation made in the welfare of intensively farmed animals (Council et al., 2009). The five freedoms are described as follows:
1. Freedom from hunger and thirst 2. Freedom from discomfort
3. Freedom from pain, injury and disease 4. Freedom to express normal behaviour
5. Freedom from fear and distress - which include the prevention of mental suffering.
To prevent suffering of any kind, it is of interest to know how broad the cognitive abilities within animals are. Perhaps their judgement of a situation can tell us more about how they feel in a certain housing or handling process.
7 Cognitive bias
Cognition, as defined by Shettleworth (2009), “the mechanisms by which animals acquire, process, store and act on information from the environment”. The mechanisms referred to consists of learning, memory, perception and the ability to make decisions. How cognitive processes can affect emotions is often explained through the term “appraisal”. Appraisal is synonymous with the term evaluation. Appraisal theory seeks to explain that our interpretation of a certain stimulus or situation can affect our emotional responses (Scherer, Schorr, and Johnstone, 2001). Which in a broad sense shows that different stimuli affect the emotional response of the individual through a cognitive decision based on how the individual evaluate the current stimulus. For example how a person can perceive, or appraise, a situation as unjust after an action that has been taken by another human being. The perceived notion about an unjust treatment can then lead to the conclusion that an aggressive approach could remove the blockage presented by the other individual. The result is the emotion anger (Grandjean and Scherer, 2008).
Corresponding behavioural and psychological effects, as well as within the nervous system, has been observed between humans and animals when exposed to situations which are highly emotion-triggering. The results has mostly been shown in more extreme cases as when a human is depressed and an animal exposed to high levels of chronic stress (I. Veissier et al., 2001). I. Veissier and colleagues (2001) showed that regrouping of calves induced an increased sensitivity for Adrenocor-ticotropic hormone (ACTH), used to regulate levels of the stress hormone cortisol. Which is similar to the result with depressed patients, where higher levels of cortisol was measured in the patients after they experienced a Developmental Screening Test (DST)-test (Carroll et al., 1981) as well as an increased activity of ACTH and higher cortisol levels after a dexamethasone/corticotropin-releasing hormone (DEX/CRH)-test (Modell et al., 1998).
If the connection between how emotions arise, as an aspect of stimulus evaluations, in humans also can be applied for other species, Mendl and colleagues (2009) proposes that we could utilise these similarities to assess indicators for emotional states. For example by presenting animals with stimuli which are associated with clear emotional states in humans (Désiré et al., 2002; Désiré, Veissier, Després, and Boissy, 2004). But that is not all, emotional states can affect cognition the other way around as well, meaning cognitive performance can be measured and utilised as indicators for different emotional states (Mendl et al., 2009).
more negatively due to their emotional state. Depressed or anxious individuals tend to have a judgemental bias toward the probability that negative events have a higher chance to occur (Butler and Mathews, 1983; Krantz and Hammen, 1979).
There are three different main groups of cognitive biases which seem to correlate with, and be the result of different emotional states; judgement-, attention- and memory biases (Paul, Harding, and Mendl, 2005). The first of which is within focus for this study.
Attention bias can roughly be explained as our perceptions being influenced by differences in our attention (Bar-Haim, Lamy, Pergamin, Bakermans-Kranenburg, and Van Ijzendoorn, 2007). Memory bias can be explained as the ability to recall certain memories based on an individual’s current emotional state (Tryon, 2014).
Let us take a deeper look into judgement cognitive bias, how it is measured and what the results indicate in the light of animal welfare.
8 Cognitive judgement bias
Judgement bias can be defined as the probability an individual assesses an outcome of a response to a ambiguous stimulus as either negative or positive (Mendl et al., 2009).
The most classic example of how a judgement bias test is executed involves a test subject which learns that two different stimuli result in a positive or negative outcome. One may have a positive result such as food reward while the other has a more negative or less favourable outcome. An ambiguous stimulus or signal is afterwards presented to the test subject who has to perform a measurable assessment of the situation, making a judgement and act on it. These intermediate stimulus occur either without or with an occasional reward (Clegg, 2018). An example of what a measurable assessment can include is found in a study on pigs where their response to the ambiguous stimulus was based on whether the pigs approached a hatch or not (Douglas et al., 2012). A more detailed description of the study can be found in section 8.2.
There are several different ways to measure the response of the animals. In Douglas and colleagues (2012) study on pigs the go no-go method was used. The go no-go method consists of teaching the animal to discriminate between two cues, a negative and a positive one. The pigs in Douglas and colleagues (2012) experiment were taught that a certain sound indicated a food reward if they approached the hatch. While the negative sound indicated that something unpleasant was about to occur, in this case the presence of a waving plastic bag in the pig’s face, if they chose to approach the hatch. In short the pigs learn when to go towards the hatch and when not to go towards it. When the ambiguous stimuli, in the form of a squeaky toy, was presented the pigs were neither offered a reward nor punishment of any kind, regardless if they approached the hatch or not. The squeaky toy was thought to be ambiguous since it was untrained but presented in the same location as the positive and negative cue. There are several other examples of when the go no-go method has been used to assess judgement bias in animals (Douglas et al., 2012; Doyle et al., 2010; Lecorps, Weary, and von Keyserlingk, 2018).
8.1 Judgement bias test in laboratory animals
The first study performed on cognitive judgement bias in animals was executed by Harding and colleagues (Harding, Paul, and Mendl, 2004), where rats were taught that a certain tone was equivalent with either a reward or a punishment. The rats learned to press a lever whenever they
to an ambiguous tone, of either 2.5, 3 or 3.5 kHz, which the rats had to assess whether it was a positive or negative. The unpredictable or less favourable environment contained element such as the cage being tilted, the bedding being damp, an unknown rat being placed in the same cage, the night/day circle could be reversed or the rat could simply be placed in an unfamiliar cage. The predictable housing was instead the same as during training sessions and did not contain any of the negative elements from the unpredictable housing. Harding and colleagues results indicate that a rat who lived in a less favourable, and more unpredictable housing environments tend to more rarely anticipate a positive outcome of the ambiguous tone, exactly like a depressed human. Other tests performed on laboratory animals include a test where they attempted to induce a cognitive judgement bias in rats. The rats were trained to discriminate between two locations, one where a food reward (positive) was placed and one were the positioned food was soaked in quinine (negative). During the training, half of the rats were exposed to a stimulus of high light levels and the other half to low light levels. When the testing started the rats were introduced to three ambiguous locations which contained neither a positive nor negative outcome (see figure 8.1). Later, half of the rats from each subgroup were exposed to a different light intensity than before, switching from high to low levels or the other way around, similarly to a crossover trial. I.e. one fourth of the rats were only exposed to high light levels through the duration of the test and one fourth were only exposed to low light levels, one fourth was first exposed to high light levels, followed by low light levels, and the final fourth was first exposed to low light levels, followed by high light levels. The results show that rats who were previously exposed to high levels ran faster to all locations when exposed to lower light levels, indicating that lower light levels were less stressful and induced a more positive emotional state than the higher levels of light. This test could be usable to address quick changes in emotional state in animals as well as to differentiate between quick and constant emotional states (Mendl et al., 2009).
Figure 8.1: Illustration of the experimental apparatus used by Mendl and colleagues (2009). A
judgement bias test which assess whether a rat’s emotional state can depend on different light levels.
An induced negative emotional state have been shown to affect the response ability in Rhesus macaques, in this case meaning that more anxious macaques acted slower. The macaques were trained to press a grey box on a touch screen, hence receiving a food reward. They were then presented with the same grey boxes both with another macaques face in them, either averted (perceived as submissive) or directly staring back (dominant). The results were compared with results gained from how the macaques performed after having gone through an unpleasant and stressful veterinary examination. In conclusion the macaques acted slower towards the direct gaze after the veterinary examination than before. The response towards the averted gaze pictures did not slow, indicating that it could be a freeze response to mildly threatening situations, which in turn could act as an indication for the animal’s emotional state (Bethell, Holmes, MacLarnon, and Semple, 2016).
8.2 Judgement bias test in farm animals
Since good health is an important aspect of good welfare, and in turn productivity, the area of how cognitive biases can indicate well-being in farm animals has been given quite a lot of attention (Douglas et al., 2012; Doyle et al., 2010; Lecorps et al., 2018). It has been demonstrated that
or no food but the introduction of a dog (negative) depending on its location. Half of the sheep were then restrained for a couple of hours, three days in a row. After the restrainment each day all of sheep were presented with buckets in ambiguous locations. The restrained sheep was more prone to approach the ambiguous buckets, when released the sheep who were previously restrained may have experienced a more positive emotional state or perhaps perceived the risks to be lesser (Doyle et al., 2010).
Pigs have been shown to express a optimistic cognitive bias when housed in a more enriched envi-ronment, indicating that improved housing and management can lead to a more positive emotional state. These results were acquired by teaching pigs a go no-go method were two different cues were taught to have a positive or negative outcome. The positive cue was in the form of a tine on a glockenspiel, the negative was a click-sound from a dog clicker. Afterwards the ambiguous stimulus was presented and the pigs response, in the form of approaching a hatch, was recorded. The pigs were then subjected to either a more enriched or barren environment for a series of weeks before the same test was concluded once again. Results indicated that pigs with a more enriched living environment tended to be more optimistic, since the pigs approached the hatch more often and faster (Douglas et al., 2012).
A study on pessimism in calves has shown that the more fearful a calf is the more pessimistic judgements it will make. The study also indicated that each individual make different interpreta-tions based on each individual’s level of fearfulness. The calves were taught through the go no-go method were the positive outcome occurred if they went to the right side of an apparatus, in the form of a bottle of milk. The negative outcome occurred if the calf instead moved to the left side were a empty bottle and an air puff to its face was presented. Calves who tended to take a longer time to approach bottles which were further away from the original reward location were deemed to be more pessimistic. The study also performed some personality tests which indicated a correlation between individual fearfulness and their pessimism in the judgement bias test (Lecorps et al., 2018).
8.3 Judgement bias in other animal groups
Some studies on cognitive bias has been performed on animals held in captivity, for example zoo animals. A study on tufted Capuchin monkeys (Cebus apella) tested the effects of grooming (natural behaviour) and social group dynamics on cognitive judgement bias, being the first study to assess the effect of social behaviour on judgement bias (Schino, Massimei, Pinzaglia, and Addessi, 2016). In contrast to laboratory and farm animals these monkeys chose themselves if they wanted to participate or not. The training phase consisted of two cups, one containing either a bigger or smaller reward (positive) in the form of 2-1 cheerios, respectively the other contained either a smaller reward or none at all (negative). A yellow and blue striped triangle was placed on the
outside of the apparatus to indicate which cup contained the greater reward (a small or a big one, see figure 8.2). The same cup always contained the same reward (big/small/none). Once the monkeys had learned that the triangle indicates where the reward was, the triangle was placed in a ambiguous position and thus not indicating which cup the monkey should choose at all nor how big the reward was. If the Capuchin chose the right cup (the one which usually contained the bigger reward) they were deemed to be optimistic, if they chose the other cup they were deemed to be pessimistic. During the ambiguous test both cups were baited with a quarter of a cheerio. These tests were performed either after or before the Capuchin had received grooming by another Capuchin, at the same time Schino and colleagues (2016) observed the social behaviour of each individual either before or after the test. The results indicated that a higher ranked Capuchin, which also received the most grooming was the most optimistic in the group. To just receive a grooming session (short term) once in while did not however affect the overall optimistic judgement of an individual (Schino et al., 2016).
Figure 8.2: Experimental apparatus presented by Schino and colleagues (2016). A cognitive
judgement bias test where the striped block were meant to indicate to a Capuchin monkey where it could find the greater reward. Image courtesy of Schino and colleagues (2016).
9 Discussion
Throughout the study we have seen that emotions can be assessed as a indicator for animal welfare. With the help of cognitive judgement bias tests, several factors such as different housing and living standards, handling and management of animals, we indicate that differences in perception and judgement between individuals can produce a direct measurement of animal welfare in our facilities. Introducing an entire new process of assessing animal welfare which can provide us with more insight than a regular health control, as well as opening up for further discussion concerning animals cognitive abilities and the effect of possible results on animal welfare applications. Harding and colleagues (2004)’s study on rats concluded that poor living standards made the rats more pessimistic, effectively indicating a connection between housing standards and state of mind in a non-human animal. The method used is simple and easy applicable to other species, making it a functional tool for assessing judgement bias in animals. Further studies have since been concluded, many scientist have utilised Harding’s method, refined it and implemented towards other species (Bethell et al., 2016; Douglas et al., 2012; Doyle et al., 2010; Lecorps et al., 2018). The go no-go method is widely used for judgement bias tests, it is easy to apply and produces a rather clear result. It is however prone to confounding effects, as described by Mendl and colleagues (2009). One being that when animals were taught to react to the negative stimulus, they should not move and that more pessimistic individuals should learn that the ambiguous stimulus contain no reward, thus faster learn to not respond to it. The problem which arises is that stressful manipulations can decrease learning capacity, consequently an enriched environment is believed to have advantageous effects on memory and learning. Making the whole idea with the learning phase of such studies a bit uncertain (Mendl et al., 2009).
Feenders and colleagues (2011) used another method to identify responses in animals in their study on birds, the novel object test. This test does not utilise a learning period, instead the animal is introduced to a novel object. The response in amount of locomotion and exploration expressed by the animal, in this case the birds, are recorded. As explained by Douglas and colleagues (2012) a cue has to be ambiguous and not novel for it to applicable to assess judgement bias, which show that a novel object test is not optimal for judgement bias tests. Novel object tests are instead often used for studies of personality in animals (Feenders et al., 2011; Lecorps et al., 2018).
The studies performed on farm animals have indicated several different properties which affect emotional state and welfare in livestock species. Douglas and colleagues (2012) showed that an enriched housing area can benefit the welfare of pigs, making them more optimistic and, accord-ing to my own interpretation, more curious - since the pigs became more prone to explore their
environment when less stressed. In relation to Douglas and colleagues (2012) findings Doyle and colleagues (2010) showed that restrained sheep tended to be more positive, or optimistic when released, indicating that freedom from restrainment can act as a means for increased welfare if implied in livestock facilities. Both studies show the importance of the environmental factor when handling animals, making judgement bias a applicable method to assess how suitable different facilities are for each non-human species respectively. The methods for judgement bias tests would however have to be designed differently depending on which non-human species the test is meant for.
Lecorps and colleagues (2018) study on calves showed an important association between animal behaviour and personality. Biases differs between each individual due to factors such as how we interpret the world around us and personal experiences. Since biases are individual it makes sense that judgement biases would have a connection to personality traits. The result that more fearful calves are more pessimistic is however slightly insufficient for me. The calves are avoiding the ambiguous location due to their fear of being blown in the face again, but is this an indication of being pessimistic or just plainly cautious? What if the calves who approached the ambiguous locations faster are not in fact more optimistic but instead more curious and adventurous? I believe further research is required to eliminate any possible third variables.
Trevarthen and colleagues (2019) highlighted some of the difficulties of experiments on cognitive bias. Several factors has to be accounted for, such as the differences between human tests and tests performed on non-human species, where Trevarthen and colleagues (2019) and colleagues suggest more research on attention in mice since tests developed for humans often have a higher focus on attention responses, like reaction time, than does animal studies. Trevarthen and colleagues (2019) recommend to further look into the behaviour of sniffing in mice, as could be used as an alternative approach. The food stimulus used in the mice experiment seemed to be unsatisfactory since the mice were more prone to react to the light, a proposed improvement was to choose two stimuli based on what awoke the same amount of response in the test subject.
Miller and colleagues (1963) showed that rhesus macaques (Macaca mulatta) could indicate emo-tional states in each other based on visual cues. This discovery could perhaps be implemented with cognitive bias in the future to test if different species can pinpoint biases in other individuals of the same species. This would however suggest an entirely new level of cognition and awareness within non-human species.
9.1 Ethical considerations
In order to test for cognitive bias, experiments have been performed in previous studies. For example, one experiment involves studies in mice. The mice were exposed to stressful situations before and during the experiment in order to measure the impact on their behaviour (Trevarthen et al., 2019). The study claims to have taken into account all regulations that apply to animal experiments. However, in accordance with Chapter 2, sections 1 of the Swedish Animal Protection Act (RSL, 2018), animals may not be subjected to any unnecessary suffering. The question will then be to determine which laws are ethnically correct or most so. Is it ever really acceptable to expose animals to any kind of experiment? Such questions open up for a more thorough discussion whether judgement bias tests could actually be hurtful towards the subjects themselves.
9.2 Social considerations
In today’s society, it has become increasingly more accepted that animals also have feelings (Isabelle Veissier, Boissy, Désiré, and Greiveldinger, 2009). If their behaviour and choices can be a product of their emotional state, it may in the future be of increasing interest to look further into the welfare of animals and what affects it. Causes of cognitive bias could be major factors that production must take into consideration when concerning the welfare of animals, for example in the slaughter industry, which is a stressful environment for many animals (Rey-Salgueiro, Martinez-Carballo, Fajardo, Chapela, Espiñeira, and Simal-Gandara, 2018). Debates on whether animal welfare can be further improved can develop and lead to changes in how we handle production of, and research on animals.
Acknowledgements
I would like to thank my supervisor Per Jensen for giving me guidance on how to get my study to include the different angles of incidence it needed as well as stopping me from getting too narrow-minded in certain areas of focus. The continuous encouragements and check ups have all been very appreciated and I am very thankful for everything.
Thanks to my examiner Matthias Laska I was able to push myself outside of my comfort zone and attempt to write a study in another language than my own, as well as pushing myself to acquire the best possible results.
Thanks to my opponents, who have given me thorough input on the structure and content of my study.
I would lastly like to thank my fellow classmates and teachers who have given me plenty of advice throughout my study as well as my entire education.
Bibliography
Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M. J., & Van Ijzendoorn, M. H. (2007). Threat-related attentional bias in anxious and nonanxious individuals: A meta-analytic study. Psychological bulletin, 133 (1), 1.
Bateson, M. (2016). Optimistic and pessimistic biases: A primer for behavioural ecologists [Behav-ioral ecology]. Current Opinion in Behav[Behav-ioral Sciences, 12, 115–121. https://doi.org/https: //doi.org/10.1016/j.cobeha.2016.09.013
Bateson, M., & Nettle, D. (2015). Development of a cognitive bias methodology for measuring low mood in chimpanzees. PeerJ, 3, e998. https://doi.org/10.7717/peerj.998
Beausoleil, N., Stratton, R., Guesgen, M., Sutherland, M., & Johnson, C. (2016). Scientific evalua-tion of animal emoevalua-tions: Brief history and recent new zealand contribuevalua-tions. The Journal
of New Zealand Studies, (22). https://doi.org/10.26686/jnzs.v0i22.3946
Bentham, J. (1823). An introduction to the principles of morals and legislation. W. Pickering. Bethell, E. J., Holmes, A., MacLarnon, A., & Semple, S. (2016). Emotion evaluation and response
slowing in a non-human primate: New directions for cognitive bias measures of animal emotion? Behavioral Sciences, 6 (1), 2.
Boleij, H., van’t Klooster, J., Lavrijsen, M., Kirchhoff, S., Arndt, S. S., & Ohl, F. (2012). A test to identify judgement bias in mice. Behavioural Brain Research, 233 (1), 45–54. https: //doi.org/https://doi.org/10.1016/j.bbr.2012.04.039
Butler, G., & Mathews, A. (1983). Cognitive processes in anxiety [Cognitions and Mood: Clinical Aspects and Applications]. Advances in Behaviour Research and Therapy, 5 (1), 51–62. https://doi.org/https://doi.org/10.1016/0146-6402(83)90015-2
Carroll, B. J., Feinberg, M., Greden, J. F., Tarika, J., Albala, A. A., Haskett, R. F., James, N. M., Kronfol, Z., Lohr, N., Steiner, M., de Vigne, J. P., & Young, E. (1981). A Specific Laboratory Test for the Diagnosis of Melancholia: Standardization, Validation, and Clinical Utility. Archives of General Psychiatry, 38 (1).
https://jamanetwork.com/journals/jamapsychiatry/articlepdf/492458/archpsyc\_38\_1\_001.pdf, 15–22. https://doi.org/10.1001/archpsyc.1981.01780260017001
Clegg, I. L. (2018). Cognitive bias in zoo animals: An optimistic outlook for welfare assessment.
Animals, 8 (7), 104.
Council, F. A. W. et al. (2009). Farm animal welfare in great britain: Past, present and future. Farm Animal Welfare Council.
Dawkins, M. S. (2000). Animal Minds and Animal Emotions1. American Zoologist, 40 (6).
https://academic.oup.com/icb/article-pdf/40/6/883/6080878/i0003-1569-040-06-0883.pdf, 883–888. https://doi.org/10.1093/icb/40.6.883
De Waal, F. B. (2011). What is an animal emotion? Annals of the New York Academy of Sciences,
1224(1), 191–206.
Désiré, L., Boissy, A., & Veissier, I. (2002). Emotions in farm animals:: A new approach to animal welfare in applied ethology. Behavioural Processes, 60 (2), 165–180. https://doi.org/https: //doi.org/10.1016/S0376-6357(02)00081-5
Désiré, L., Veissier, I., Després, G., & Boissy, A. (2004). On the way to assess emotions in animals: Do lambs (ovis aries) evaluate an event through its suddenness, novelty, or unpredictabil-ity? Journal of Comparative Psychology, 118 (4), 363.
Douglas, C., Bateson, M., Walsh, C., Bédué, A., & Edwards, S. A. (2012). Environmental en-richment induces optimistic cognitive biases in pigs. Applied Animal Behaviour Science,
139(1), 65–73. https://doi.org/https://doi.org/10.1016/j.applanim.2012.02.018
Doyle, R. E., Fisher, A. D., Hinch, G. N., Boissy, A., & Lee, C. (2010). Release from restraint generates a positive judgement bias in sheep. Applied Animal Behaviour Science, 122 (1), 28–34. https://doi.org/https://doi.org/10.1016/j.applanim.2009.11.003
Feenders, G., Klaus, K., & Bateson, M. (2011). Fear and exploration in european starlings (sturnus vulgaris): A comparison of hand-reared and wild-caught birds. PLoS One, 6 (4).
Grandjean, D., & Scherer, K. R. (2008). Unpacking the cognitive architecture of emotion processes.
Emotion, 8 (3), 341.
Haga, H. A., & Ranheim, B. (2005). Castration of piglets: The analgesic effects of intratesticular and intrafunicular lidocaine injection. Veterinary Anaesthesia and Analgesia, 32 (1), 1–9. https://doi.org/https://doi.org/10.1111/j.1467-2995.2004.00225.x
Harding, E. J., Paul, E. S., & Mendl, M. (2004). Cognitive bias and affective state. Nature,
427(6972), 312–312.
Haselton, M. G., Nettle, D., & Andrews, P. W. (2005). The evolution of cognitive bias (D. M. Buss, Ed.). In D. M. Buss (Ed.), The handbook of evolutionary psychology. Wiley Online Library.
Keeling, L., & Jensen, P. (2009). Abnormal behaviour, stress and welfare (P. Jensen, Ed.). In P. Jensen (Ed.), The ethology of domestic animals: An introductory text. Cabi.
Krantz, S., & Hammen, C. L. (1979). Assessment of cognitive bias in depression. Journal of
Ab-normal Psychology, 88 (6), 611.
Lecorps, B., Weary, D. M., & von Keyserlingk, M. A. (2018). Pessimism and fearfulness in dairy calves. Scientific reports, 8 (1), 1–9.
Mendl, M., Burman, O. H., Parker, R. M., & Paul, E. S. (2009). Cognitive bias as an indicator of animal emotion and welfare: Emerging evidence and underlying mechanisms [Special
Merskey, H. (1979). Pain terms: A list with definitions and notes on usage. recommended by the international association for the study of pain. Pain, 6 (3), 249–252.
Miller, R. E., Banks Jr, J. H., & Ogawa, N. (1963). Role of facial expression in" cooperative-avoidance conditioning" in monkeys. The Journal of Abnormal and Social Psychology,
67(1), 24.
Modell, S., Lauer, C. J., Schreiber, W., Huber, J., Krieg, J.-C., & Holsboer, F. (1998). Hormonal response pattern in the combined dex-crh test is stable over time in subjects at high familial risk for affective disorders. Neuropsychopharmacology, 18 (4), 253–262. https://doi.org/ https://doi.org/10.1016/S0893-133X(97)00144-9
Oatley, K., Keltner, D., & Jenkins, J. M. (2006). Understanding emotions. Blackwell publishing. Paul, E. S., Harding, E. J., & Mendl, M. (2005). Measuring emotional processes in animals: The
utility of a cognitive approach. Neuroscience & Biobehavioral Reviews, 29 (3), 469–491. https://doi.org/https://doi.org/10.1016/j.neubiorev.2005.01.002
Rey-Salgueiro, L., Martinez-Carballo, E., Fajardo, P., Chapela, M., Espiñeira, M., & Simal-Gandara, J. (2018). Meat quality in relation to swine well-being after transport and during lairage at the slaughterhouse. Meat Science, 142, 38–43. https://doi.org/https://doi.org/10.1016/ j.meatsci.2018.04.005
Rollin, B. E. (1989). The unheeded cry: Animal consciousness, animal pain and science. Oxford University Press.
RSL, N. (2018). Djurskyddslag.
Scherer, K. R., Schorr, A., & Johnstone, T. (2001). Appraisal processes in emotion: Theory,
meth-ods, research. Oxford University Press.
Schino, G., Massimei, R., Pinzaglia, M., & Addessi, E. (2016). Grooming, social rank and ‘optimism’ in tufted capuchin monkeys: A study of judgement bias. Animal Behaviour, 119, 11–16. https://doi.org/https://doi.org/10.1016/j.anbehav.2016.06.017
Shettleworth, S. J. (2009). Cognition, evolution, and behavior. Oxford university press.
Trevarthen, A. C., Kappel, S., Roberts, C., Finnegan, E. M., Paul, E. S., Planas-Sitjà, I., Mendl, M. T., & Fureix, C. (2019). Measuring affect-related cognitive bias: Do mice in opposite affective states react differently to negative and positive stimuli? PloS one, 14 (12). Tryon, W. W. (2014). Chapter 11 - clinical implications of network principles 3–12 (W. W. Tryon,
Ed.). In W. W. Tryon (Ed.), Cognitive neuroscience and psychotherapy. San Diego, Aca-demic Press. https://doi.org/https://doi.org/10.1016/B978-0-12-420071-5.00011-9 Veissier, I. [I.], Boissy, A., dePassillé, A. M., Rushen, J., van Reenen, C. G., Roussel, S., Andanson,
S., & Pradel, P. (2001). Calves’ responses to repeated social regrouping and relocation1.
Journal of Animal Science, 79 (10).
https://academic.oup.com/jas/article-pdf/79/10/2580/22142090/2580.pdf, 2580–2593. https: //doi.org/10.2527/2001.79102580x
Veissier, I. [Isabelle], Boissy, A., Désiré, L., & Greiveldinger, L. (2009). Animals’ emotions: Studies in sheep using appraisal theories. animal welfare, 18.
Xie, W. (2011). Assessment of pain in animals. In C. Ma & J.-M. Zhang (Eds.), Animal models of
pain (pp. 1–21). Totowa, NJ, Humana Press.
