1-1 Introduction to Biological Psychology

By the second half of the twentieth century, psychologists interested in human behaviour were becoming increasingly dissatisfied with the limitations of the behaviorist approach. Psychologists working on animal learning also realised that some of the phenomena that could be observed in their experiments could only be explained by postulating intervening cognitive mechanisms. Tony Dickinson, in his short text Contemporary Animal Learning Theory, published in 1980, used the example of sensory preconditioning. Dickinson was one of the initiators of the so-called ‘cognitive revolution’ in animal learning.

In a standard Pavlovian procedure, rats are initially trained to press a bar for small pellets of sweetened food. Then they are exposed to an initially neutral stimulus – a light in the wall of their cage – immediately prior to receiving a mild foot shock. After several such pairings the rats are exposed to the light alone – there is no shock. Nevertheless, the rats ‘freeze’ (remain immobile) and pause bar-pressing for the food reward for a few seconds before returning to their normal behaviour.

The critical modification in sensory preconditioning is to expose the rats to pairings of a light with a second neutral stimulus, in this case a sound, prior to the main conditioning task. Since nothing happened after these initial pairings the rats rapidly come to ignore them. Their ongoing behaviour, in this case bar-pressing for the food pellets, was not changed. Now the rats were conditioned to associate the light and shock in the standard way. Following conditioning they were exposed either to the light, or to the tone. Rats exposed to the tone paused in a similar way to rats that were exposed to the light despite never having experienced that sound preceding a shock. So, despite little or no change in their behaviour during those initial light-tone pairings, they clearly had learned something. Learning does not have to involve any overt behavioural change.

Dickinson set out the implications in the following way:

As we shall see, sensory preconditioning is but one of many examples of behaviourally silent learning, all of which provide difficulties for any view that equates learning with a change in behaviour. Something must change during learning and I shall argue that this change is best characterised as a modification of some internal structure. Whether or not we shall be able at some time to identify the neurophysiological substrate of these cognitive structures is an open question. It is clear, however, that we cannot do so at present. (Dickinson, 1980, p.5)

That was 1980!

As I write this, in the 2020s, it is possible to give an optimistic answer to Dickinson’s final query. Modern techniques from neuroscience can identify the brain structures and changes in small groups of neurons that support learning and memory. In one recent example, Eisuke Koya and his colleagues, who work in the School of Psychology at the University of Sussex, describe a simple learning task in which mice are exposed to a short series of clicks (the ‘to-be-conditioned’ stimulus) followed by an opportunity to drink a small quantity of a sucrose solution. After being exposed to such pairings over a period of a few days, they learn to approach the sucrose delivery port as soon as the clicks begin. The mice used in these experiments were genetically modified in such a way that nerve cells in the frontal cortex that were activated during the conditioning trials could be made to glow with a green fluorescence. The research established that small, stable groups of nerve cells (‘neuronal ensembles’) were activated from one day to the next. The researchers were also able to manipulate these cells in such as way as to produce an abnormal change in their activity during subsequent tests in which the animals were exposed to the conditioned stimulus. Approach to the sucrose delivery port was disrupted when this was done, suggesting that the activation of these cells contributed in an important way to the learnt behaviour of the mouse (Brebner et al., 2020). Experiments of this kind are approaching the goal of identifying the neural structures and mechanisms that support learning and memory, and demonstrate how psychologists and neuroscientists can collaborate to tackle the fundamental problems of biological psychology.

Field studies of non-human primates

Researchers in the field of animal learning were not the only ones who found it difficult to explain their experimental results without invoking cognitive processes that could not be directly observed. Ethologists faced a similar challenge.

Jane Goodall began her fieldwork on chimpanzees in the early 1960s and, as she got to know and was accepted by the troop that she was studying at Gombe Stream in Tanzania, gathered evidence about their rich social and emotional lives and the use of tools that revolutionised studies of primate behaviour. To the consternation of her colleagues at Cambridge, she also gave names to the chimpanzees rather than the numbers which would supposedly lead to more objective study.

At about the same time, Alison Jolly was beginning her studies of lemur behaviour in Madagascar. She argued that the major driving force in evolution of primate cognition came from the complex demands that come from living in complex and long lasting social groups (Jolly, 1966).

Field studies carried out since then have demonstrated long-lasting social relationships in primate species such as baboons and vervet monkeys. Calls that the animals make in the context of aggressive encounters are interpreted in terms of an animal’s prior knowledge of social hierarchies within their group and their prior behaviour. For example, female chacma baboons, studied by Dorothy Cheney and Robert Seyfarth, have a call, the ‘reconciliatory grunt’, that is given just after an aggressive encounter to indicate a peaceful conclusion between the two individuals. When the call was played to a female who had just been involved in a mutual grooming session with another female, she behaved in a way that implied that the call must be directed at someone else and not her. However, if she had been involved in an aggressive encounter with that same female a little earlier, she behaved in a way that implied that the grunt had been directed at her (Engh et al., 2006).

Field studies of other long lived mammals, including elephants and dolphins, suggest that they also have complex social networks and sophisticated cognitive abilities.

Two features stand out at the end of this very short historical survey of ideas about the relationship between the brain and behaviour. The first is that, among contemporary psychologists and neuroscientists, there is an almost universal acceptance of some form of materialism. In other words all of the complexity in our behaviour, including such relatively less well understood areas such as consciousness, are a consequence of physical mechanisms operating in our bodies, and primarily in the nervous system. The second is that, despite a hiatus that lasted for at least the first half of the twentieth century, the study of phenomena such as emotion and consciousness are no longer seen as ‘off limits’ for scientific study. One challenge for modern neuroscience is to understand how the nervous system builds, uses and attaches emotional weight to internal representations of aspects of the external world.

Since biological psychology is concerned with both behaviour and relevant physiological and brain mechanisms, it will often start with some interesting behavioural observations or experimental data. Once the behaviour of interest has been adequately documented, it is time to ask some questions. The ethologist Niko Tinbergen suggested that were four broad kinds of questions that might be of interest (Tinbergen 1963). The first two have a timescale within the life cycle of an individual animal and are concerned with:

(i) the underlying causes of changes in behaviour, such as brain mechanisms or hormonal changes, and

(ii) the development of behaviour, for example as an individual matures to adulthood.

These are sometimes referred to as the proximate causes of the behaviour. The remaining two questions are set in a much broader time frame and can be thought of as the ultimate causes (Bateson & Laland, 2013). They are concerned with:

(iii) the evolutionary relationships between patterns of behaviour in different species, and

(iv) the advantages of particular patterns of behaviour in the context of natural selection.

A couple of examples should illustrate how these questions differ from one another and yet still address the question of why a particular behaviour occurs in the form that it does.

Example 1: Bird song

The chaffinch (Fringilla coelebs) is a common European songbird. In the early Spring, adult males have a striking breeding plumage and their bills darken. At the same time the birds begin to perch in conspicuous places within a small territory that they defend from other males and sing. The females do not sing, although they use a variety of other calls. So, why do male chaffinches sing?

Causes and mechanisms. The changes in bill colour and onset of singing are associated with the increase in day length in the Spring. Experimental studies in other songbird species, such as those by Fenando Nottebohm in canaries, have shown that there is an increase in the secretion of testosterone at this time, and that experimental administration of this steroid hormone produces the same changes in appearance and behaviour (Nottebohm, 2002). Further studies have demonstrated that there are changes in the bird’s brain at this time.

The most surprising result was a clear demonstration that new neurons are formed through cell division in the areas of the brain involved in song. In the mid 1980s, when these studies were performed, the consensus was that neurons were never added to a mature vertebrate brain. Nottebohm’s work forced a re-examination of this idea. The methods that his lab used were repeated in other species, and demonstrated that the same thing could also happen in mammals, including humans. In the years since these ground-breaking studies, a good deal more has been learned about circuits within the songbird brain that support song production.

Development. You do not need to be an experienced birdwatcher to recognise the song of a chaffinch. It consists of several short trills followed a characteristic terminal flourish. This chaffinch sonagram shows the loudness (upper panel) and frequency changes in the song of a chaffinch. Alternatively, play the song as a short video.

However, although it is difficult to mistake the song of a chaffinch for that of a different species, the song has unexpected complexity. An individual may sing several subtly different song types and there are also clear differences in the song over the different parts of their widespread distribution through Europe. This raises several questions about the way in which a young chaffinch develops its song repertoire. If a nestling is reared in an acoustically isolated environment, it develops a highly abnormal song lacking the detailed structure typical of a chaffinch. However, if a bird reared in the same way is exposed to tape-recorded song, then it accurately copies the song types and incorporates them into its repertoire. Of course there will be be a considerable gap between hearing the song as a summer fledgling, and then singing the song in the following spring, so this is another example of behaviourally silent learning. In the wild, things turn out to be be much more complex. A chaffinch may acquire some song types in the first summer, but additional ones may also be learnt from neighbours in the following spring as they set up territories. It is clear that one early, popular idea about the development of song types is incorrect: they are not exclusively learnt from a bird’s father (Riebel et al., 2015).

Evolutionary relationships. Birds, like mammals, reptiles and amphibians, are vertebrates. Although there is tremendous variety in their appearance and behaviour, there are common features, such as the presence of a backbone. There are also strong similarities in broad aspects of brain structure and functioning. Songbirds are one of several groups of living birds and there are some 5000 different species. They all have a well-developed syrinx (the rough equivalent of the human larynx) which has a complex musculature that allows the bird to sing. There is tremendous variation from one species to another, from the complex, extended song of thrushes such as the blackbird and nightingale, to the much simpler song of the chaffinch. Songbirds evolved about 45 million years ago, and birds diverged from the vertebrate line that also gave rise to mammals some 320 million years ago. Our common ancestor probably resembled a small lizard. Present day lizards can show some degree of behavioural flexibility and social learning, so there are interesting questions about the extent to which some of the more advanced cognitive abilities of birds and mammals evolved independently or build on components already present in that common ancestor.

Function or current utility. Song is energetically expensive at a time of the year when food is not at its most abundant. It can also be dangerous. There is a risk that a sparrowhawk will appear over the top of the hedge on which the chaffinch is singing and provide the hawk with its next meal! It follows that song must also potentially enhance the biological fitness of the bird in some way. There are at least two factors at play here. A male has to attract a female to nest in his territory and mate with her. His song also advertises to other males that he holds, or is attempting to hold, that space, and may be a prelude for fighting over ownership of the best areas. There is also experimental evidence that characteristics of the song, particularly the complexity of the final trill, may be attractive to females and lead them to prefer one male over another. Although the evidence is not completely convincing for chaffinches, this idea would fit in with findings from a variety of other species. The croak of a frog, the roar of a red deer stag or the colours of a peacock’s tail may act as signals about the quality of the individual making the call, and may have the advantage of being hard to falsify – so called ‘honest signals’. However the precise way in which such signals evolve remains unclear (Penn & Számadó, 2020; Smith, 1991).

Example 2: Anxiety and fear

Tinbergen’s general approach can be productive in thinking about any aspect of behaviour. In humans anxiety or fear is an unpleasant emotional experience that may come in many forms including panic and phobias of various kinds. The emotion of fear is often evoked by quite specific threat-related stimuli – perhaps a snake (snake phobia) or wide open spaces (agoraphobia). In the same way as for bird song, we can ask the question ‘why?’ and break it down into queries about either proximate or ultimate causes.

Causes and mechanisms. The underlying physiological and brain mechanisms are well studied. They include increases in heart rate, release of hormones such as adrenalin, and activation of a specific brain network that includes the amygdala. One part of the amygdala, the central nucleus, is responsible for activating these different physiological changes in a coordinated manner (LeDoux, 2012). An understanding of these types of mechanisms has clinical relevance. Drugs that act selectively on these threat-processing circuits may have value as treatments for anxiety. Indeed benzodiazepines such as valium  are still widely used in this way and are known to have especially potent effects in the amygdala. An important, but still unanswered, question is how these physiological responses relate to the conscious feeling of fear.

Development. We also know a good deal about the way in which fear may develop during an individual’s lifespan. Simple conditioning may often play a role and there is also good evidence that species as varied as rodents and primates are more likely to become fearful and avoid some types of object rather than others. In social species observational learning may also be important. Studies of young rhesus monkeys illustrate these points. A rhesus infant will initially show little avoidance or fear of model snakes or flowers. However if they are allowed to watch an edited video in which an adult rhesus appears to respond fearfully to either the flowers or a snake, they themselves develop fear responses to the snake but not to the flower (Cook & Mineka, 1990). This suggests an innate tendency to become fearful of some kinds of object, such as a snake, that can potentiate the effects of observational learning. In a similar way many rodent species will avoid odours associated with potential predators such as a fox or cat without having any previous experience of those animals. However, especially when young, those responses may be amplified if they observe an adult responding strongly to the same stimulus.

Evolutionary relationships. Comparative studies of the specific behaviour patterns associated with fear and the underlying physiological and brain mechanisms suggest that they have been conserved through vertebrate evolution. Charles Darwin, in the Expression of the Emotions in Man and Animals, provided some of the first really detailed behavioural descriptions of facial expressions associated with fear and especially emphasised their role in communication. Detailed comparisons of the neural circuitry in mammals, birds, amphibians and reptiles suggest that the amygdala, and especially its connections to the autonomic nervous system which activate the hormonal and other physiological responses to fear-evoking stimuli, are conserved through the entire vertebrate evolutionary line and must therefore have originated at least 400 million years ago. So it is not surprising that one of the responses of a Fijian ground frog to the presence of a potential predator (a cane toad) is an increase in the stress hormone corticosterone as well as a behavioural response, in this case immobility, that reduces the likelihood of being eaten (Narayan et al, 2013). Exposure to a stressful situation in humans produces the same hormonal response, although it is cortisol, which is almost structurally identical to corticosterone, that is released.

Function or current utility. Questions of function, or current utility, can be thought about at multiple levels. It is clear that fear, or the perception of threat-related stimuli can be a powerful driver of learning. As we saw earlier in this chapter, previously neutral stimuli that predict threat or danger come to evoke the same responses as the threat itself (Pavlovian conditioning). In the natural environment such responses are likely to enhance biological fitness. However in addition to thinking about the likely function of fear systems in a rather global manner, it is also possible to analyse the individual behavioural elements that make up a a fear response. One such element, described by Darwin (Darwin, Charles, 1872) and also recognised in the later studies of Paul Ekman, is that the eyebrows are raised which results in the sclera (white) of the eye becoming much more obvious (Jack et al., 2014 includes a video example). The original function of this response may simply have been to widen the field of view but, especially in primates, it can now also serve as a way of communicating fear within a social group. It is likely that behavioural responses frequently gain additional functions during evolution, perhaps even making the original function irrelevant. This is the reason that the term ‘current utility’ is often preferred to function (Bateson & Laland, 2013). If a functional hypothesis is be tested experimentally it will always be current utility that will be assessed. When a particular characteristic or feature acquires additional functions in this way they are sometimes described as exaptations rather than adaptations.

The first phase of any scientific investigation is likely to be descriptive. In biological psychology, this is a point at which the influence of an ethological approach is most obvious. It is easiest to describe how this phase proceeds by using some specific examples. Once the behaviour of interest has been clearly characterised it is often time to collect some empirical data. This will often involve either collecting behavioural and physiological data and correlating them, or taking a more experimental approach in which environmental or physiological factors are deliberately manipulated. A combination of these approaches will begin to elucidate the way in which neural processes influence behaviour and, in turn, are influenced by the consequences of that behaviour.

Describing behaviour: facial expressions and individual variation during conditioning

Many mammals, including rodents and primates, (including humans in the latter category), make distinctive facial expressions as they try out potential food sources. Humans will lick their lips as they eat something sweet and gooey. A food or drink that is unexpectedly sour (like pure lemon juice) or bitter (perhaps mature leaves of kale or some other member of the cabbage family) might elicit a gaping response in which the mouth opens wide and, in more extreme cases, saliva may drip out of the mouth.

These kinds of response can be observed in quite young babies. Indeed, as any parent is likely to know, they are very common as an infant transitions from breast feeding to solid foods. It may seem surprising but very similar responses can be observed in rats or mice as they drink sweet, sour or bitter solutions. It demonstrates that these are responses that are likely to have been conserved over relatively long periods of evolutionary time. They may serve a dual function. A response like gaping will help to remove something that may be toxic from the mouth – bitterness is often a signal that a plant contains harmful toxins. However it is also likely that, at least in some species, the ‘current utility’ of these expressions also includes a communicative function in species that feed in social groups. This would be another example of an exaptation (i.e. an additional function that becomes adaptive later).

Detailed measurement of these facial responses forms the basis of the so-called ‘taste reactivity task’ in which controlled amounts of solutions with different taste properties are infused into the mouth of a rat or mouse and the facial expressions quantified (Berridge, 2019). The task was initially devised to investigate the role of different brain structures in taste processing. A little later some detailed studies with a variety of different solutions revealed that the extent to which they evoked ingestive (‘nice’) and aversive (‘nasty’) responses could, at least to some extent, vary independently. It then became clear that there were drug and brain manipulations that had no effect on the facial expressions evoked by a liked, or rewarding, sweet solution. However those same manipulations did reduce the extent to which an animal would be prepared to work (e.g. press a lever, perhaps several times) in order to gain access to that solution. The important implication was that the extent to which something is ‘liked’, measured by facial expressions, may depend on different factors to those that affect whether it is ‘wanted’, measured by effort to obtain that reward. Although this distinction first arose in the investigation of what might seem an obscure corner of biological psychology, it has, as you will read in Chapter 15, Addiction, become an important theoretical idea that helps explain some otherwise puzzling features of addiction to drugs, food and other rewarding stimuli.

One feature of animal behaviour, humans included, that seem irritating at first is that there is often substantial variation between individuals exposed to the same experimental manipulation. It can be tempting to ignore it, or to choose measures that at least minimise it. But this can be a mistake, as this example from Pavlovian conditioning demonstrates.

As a group of rats learns the association between the illumination of light and the delivery of a food pellet they retrieve the pellet more rapidly and the averaged behaviour of the group generates a smooth ‘learning’ curve. However, careful observation of individuals reveals several things. First, the changes in individual animals are much more discontinuous – almost as though at some point, individual rats ‘get’ the task, but at different times during the training process! The behaviour of individuals also varies in other, potentially interesting, ways. Some rats approach the light when it illuminates, rearing up to investigate it, and continue to do so even when they subsequently approach the place where the pellet is delivered. Other rats move immediately to the location where the food pellet will be delivered, apparently taking no further interest in the light. The behaviour of the former group is referred to as ‘sign tracking’ and the latter as ‘goal tracking’. It turns out that sign trackers and goal trackers differ in other interesting ways. For example, sign-tracking rats show greater impulsivity in other tasks and acquire alcohol self administration more readily. The same kinds of distinction may also show up in human behaviour and predict vulnerability to drug addiction and relapse.

Although the descriptive phase is likely to begin any serious scientific investigation, once interesting observations have been made they are likely to raise the kinds of questions that were discussed in the last section. What are the brain and physiological mechanisms that underlie the behaviour? How do the behaviour patterns develop through lifespan? and so forth. In working out how to tackle such questions there are a number of potential ways forward. They typically use a combination of two general strategies.

Investigational strategies: correlational approaches, and experimental manipulations

One strategy is to observe changes in behaviour, typically in a carefully controlled test situation or using a clearly defined set of behaviour patterns when doing fieldwork, while measuring changes in physiological and brain function that are likely to be relevant. Then, using appropriate statistical techniques, the changes in behaviour and physiology can be correlated together. The second strategy is to deliberately manipulate a test situation in order to determine the extent to which an imposed change in in physiology or brain function leads to a change in behaviour, or a change in behaviours leads to a physiological change.

The issue with the first correlational strategy is that, although the results may suggest that there is some type of causal relationship between behaviour and physiology, they don’t clarify the nature of the relationship. Behaviour A may cause changes in physiological or brain variable B. But equally, the physiology may influence the behaviour. Finally, it may be that there is no direct mechanistic linkage between behaviour and physiology. Instead, some third variable is independently affecting both. A further complication is that there may be important feedback loops that influence the outcome. So, although correlational studies can be very useful, they do have limitations when it comes to their interpretation. The second strategy, involving deliberate experimental manipulation, has a better chance of determining the direction of causation. But it may raise other problems. Deliberately interfering with the functioning of a complex biological system may lead it to respond in unpredictable ways and may also be ethically problematic.

Let’s see how this works in practice by considering the behaviour of red deer during the autumn rutting season. At this time, a successful male deer may establish a ‘harem’ of female deer and defend them against other males. His behaviour makes it more likely that only he will have the opportunity to mate with them, and hence that the resulting offspring would enhance the representation of his genes in the next generation. Females also gain from being in the harem by gaining some protection from harassment by other males, which provides the opportunity to feed in a more uninterrupted way. They may choose the harem of a particular male on the basis of his perceived fitness.

It would first be possible to correlate the gradual increase in aggressive behaviour during the red deer rut with the increasing blood testosterone that occurs during the autumn. It might be tempting to conclude that the increased testosterone level causes the increased level of aggressive behaviour directed at other males. However there are at least two, and perhaps more, possibilities that would need to be excluded first.

First, it is possible that the changes in behaviour and increased testosterone level are triggered independently by some third factor. One alternative would be that decreasing day length acts as that trigger. The day length signal might be detected within the pineal gland (Descartes’ supposed valve from the soul to the body!) and independently trigger both the hormonal and behavioural changes.

A second possibility is that the behavioural changes actually trigger the hormonal change. This isn’t a completely implausible suggestion. Such effects have been documented in a number of mammals, including human (male) tennis players, where testosterone levels increase after a match that they have just won. In the same way, testosterone secretion might be sensitive to whether aggressive encounters between the male deer are won or lost.

An experimental approach can be used to overcome the difficulty in deciding what causes what. In the case of the role of testosterone in aggression, one possible strategy is to remove the source of testosterone and determine whether aggressive behaviour continues. In fact this has been common practice for millennia in managing farmed animals. An intact bull may be very aggressive but castrated male cattle (bullocks) are typically much less so. In experimental animals the specific role of testosterone could be demonstrated by administering the hormone to a castrated animal and showing that aggressive behaviour returns to the expected level. Experiments of exactly this kind have been performed on red deer, and indicate that testosterone does indeed restore rutting behaviour when administered to a castrated male in the autumn. However the effect of the hormone on rutting behaviour is absent when the same treatment is given in the spring, although there is some increase in aggressive behaviour (Lincoln et al., 1972). So, factors like day length and hormone level interact in a more complex way than might be expected.

A similar experimental approach can be taken in relation to the contribution that particular brain structures or identified groups of neurons may make to a specific behaviour pattern. Suppose that we have already discovered, perhaps by making recordings of neural activity, that neurons in a particular structure (let’s call it area ‘X’) become more active while an animal is feeding. How can we demonstrate that those neurons are important in actually generating the behaviour rather than, for example, just responding to consequences of eating food? In other words, how can we determine that activation of area X causes feeding as opposed to feeding causing activation of area X? If stimulation of the nerve cells within area X leads to the animal, when not hungry, beginning to feed on a highly palatable food, then you might assume that demonstrates their critical role – authors reporting such an experiment will often state that the cells in this area are ‘sufficient’ to generate feeding. However this doesn’t show that those same cells are always active in the many other experimental situations in which that animal might begin to feed. In the same way, suppose an animal fails to eat when the cells in that same area X are inactivated. That finding doesn’t demonstrate that these cells are always ‘necessary’ for feeding to occur. For example, suppose the original test situation were eating a palatable food when already sated, the animal might still eat when those same cells were inactivated after they had been deprived of food for a few hours (Yoshihara & Yoshihara 2018). It is also possible that inactivation of those cells might interfere with other types of behaviour, suggesting that they had no unique importance in relation to feeding. This highlights the moral that demonstrating causation, and especially the direction of causation, is rarely easy!

In studies where only correlations have been measured it is tempting for those presenting the research to slip from initially saying that an association has been observed to then discussing the results in terms of a causal mechanism that hasn’t been fully demonstrated. This is something to watch for when critically assessing research publications. It often happens in the discussion of results based on fieldwork when an experimental approach may be much more challenging. A combination of correlational and experimental approaches can also be taken in studying questions about the development of behaviour.

This combination of approaches can be taken in the study of human behaviour and its relationship to particular brain or other physiological changes. However, the experimental approach can be limited by the more significant ethical concerns that may arise. Clinical data has been used since the time of Galen, Ibn Sina, Broca and others to correlate the brain damage that occurs after strokes, or other forms of brain damage, with changes in behaviour. Classic studies include the patient Tan, studied by Broca, where loss of the ability to speak was linked to damage in the left temporal cortex; and Phineas Gage, whose damage to the prefrontal cortex was associated with more widespread changes in behaviour. This last example also provides a cautionary tale. There is an unresolved dispute as to how substantial and permanent Gage’s changes in behaviour actually were, despite the typical clarity of textbook presentations (Macmillan, 2000).

Ethical issues in Biological Psychology