Philosophy of the Social Sciences essay

Tue, 24 Apr 2018

This is an essay submitted for Philosophy of the Social Sciences, as part of my MSc in Philosophy of the Social Sciences.

According to some philosophers, mental states are holistic. When we rationalize action, inferring back from people’s actions to their beliefs and desires requires knowing more about further beliefs and desires. Donald Davidson captures the point in the following words: “When we attribute a belief, a desire, a goal, an intention or a meaning to an agent, we necessarily operate within a system of concepts in part determined by the structure of beliefs and desires of the agent himself. Short of changing the subject, we cannot escape this feature of the psychological; but this feature has not counterpart in the world of physics” (2001: 3). Why does Davidson think that the holism of the mental makes the prospect of understanding people scientifically – the prospect of a science of the social - so difficult? Discuss and critique.

  1. Introduction
The natural sciences have experienced considerable success in formulating laws that explain and predict the natural world.  The social sciences, in contrast, have struggled to formulate laws that explain and predict human behaviour. In this essay, I aim to consider an important source of difficulty facing the social sciences: what Donald Davidson has labelled “holism of the mental”.  This is the idea that mental states, such as beliefs and desires, with which we usually explain human behaviour, are not directly observable. They must instead be logically implied from actions, and this process depends on assumptions, for example other beliefs and desires that the individual holds.  This leaving the whole endeavour susceptible to error.

Davidson has not written specifically or extensively on holism of the mental, mentioning it within his essays Mental Events (1970) and Psychology as Philosophy (1974) as part of arguments on the impossibility of psychophysical laws; that is, laws in which mental states cause or entail physical states (Davidson, 2001).  Holism of the mental should also not be confused with the parallel idea of semantic holism: that the intended meaning of any term depends on the meanings of all other terms (Jackman, 2017).  As a result, I have focussed on what I understand to be the likely implications of holism of the mental, drawing in part from discussion of mental explanations in Rosenberg (2015). While I will accept that holism of the mental poses difficulties, I will offer some ways in which they may, to some extent, be overcome.  

The essay will be structured as follows: In section 2, I describe Davidson’s view of the role of mental explanations of behaviour, and indicate how the differ from physical explanations.  In section 3, I acknowledge the difficulties posed by holism of the mental. In section 4, I suggest some ways in which progress might nonetheless possible.

  1. Mental and physical explanations of behaviour
Davidson’s consideration of laws within the social sciences relies upon the distinction between two kinds of explanations of human behaviour: mental (psychological) and physical.  Mental explanations are the ones most often used, and I will first describe these and explain how Davidson understands these working. I will then describe physical explanations of human behaviour, highlighting how they differ to mental explanations.

Our usual way of explaining human behaviour is with mental or psychological explanations.  These are used intuitively by most people, hence are sometimes referred to as “folk psychology”.  Mental explanations are sets of beliefs and desires (or more generally any such propositional attitudes), that, combined with other beliefs and desires, rationally justify the observed behaviour.  For example, we say that a man carried an umbrella because he believed it might rain and desired not to get wet. Not all behaviour can be rationalised this way, for example mere (unintentional) behaviour, but we designate as action that behaviour we can explain by beliefs and desires.  Mental explanations are subjective (dependent on how we interpret the action) and implied rather than objectively observed.  

Physical explanations are of the type more typical in natural sciences, and involve identifying objectively observed initial conditions, and laws (perhaps but not necessarily physical mechanisms) that determine the resulting outcome.  In the case of human behaviour, we may expect it to result from an initial physiological state (including the state of the brain), via physiological mechanisms. Unfortunately, explanations of this type have been of limited use within social science, as behaviour depends greatly on variation in the state of the brain, and this cannot currently be known precisely enough to form accurate predictions.  

Davidson believes that mental explanations are the appropriate ones for human action (2001, p239); that psychology should take philosophy as its template instead of the natural sciences like neuroscience or even physics.  However, as Davidson recognised, mental explanations have limitations, which I will consider in the next section.

  1. Difficulties posed by mental explanations
In the previous section I described mental explanations, which Davidson believes are the appropriate ones for explaining human behaviour.  However, mental explanations, though widely used, have struggled to provide law-like generalisations of the sort that have contributed to the success of the natural sciences.  I argue that much of this difficulty is due to the holism of the mental: the fact that beliefs and desires do not operate individually, but collectively, as a whole. This section will discuss the difficulties that holism poses to mental explanations. Firstly, I describe how individual beliefs and desires depend on the other beliefs and desires for their significance.  Secondly, I describe how this allows indeterminism: multiple sets of beliefs and desires can rationalise observed actions, which makes it difficult to choose between them. Finally, I discuss how unconsidered factors, can undermine the logical relationship between beliefs and desires and behaviour.

According to Davidson’s understanding, mental explanations consist of beliefs and desires logically implied from an individual’s behaviour.  This logical process of implication requires us to assume many of their other beliefs and desires. For example, if we see someone choosing chicken from a menu, we would likely imply that they desire chicken, but this logic relies on them being able to read the menu and believing that the item was in fact chicken.  We may practically consider one belief or desire to be the primary reason for an action, but this is only because we have taken for granted other beliefs or desires, and the lack of further beliefs and desires. Holism of the mental means that no belief or desire can be deduced with certainty, each requires the assumption of others.  Furthermore, we can never rule out a belief or desire: we can always explain additional actions by supposing additional beliefs and desires.

Because actions are more correctly explained by an individual’s full set of beliefs and desires, we can often explain action in multiple ways, and our choice cannot be objective.  For example, on seeing a man picking up someone else’s umbrella, one viewer might explain it as him mistakenly thinking it is his own, while another might explain it as him intentionally stealing.  One could interpret this indeterminism as suggesting that no set of beliefs and desires can ever be considered the correct one. I dispute this point; there exist some beliefs and desires that we would maintain even if we had all possible evidence of the individual.  However, we must accept that more evidence may change our view of which alternative is most plausible, and the difficulty that this uncertainty adds to our being sure of our beliefs and desires.

Finally, other factors, beyond beliefs and desires, can have an influence on our behaviour, blurring the scope of mental explanation.   Davidson refers to this problem as that of the mental not being a closed system, as it is affected by non-mental factors (2001, p224). For example, I may believe that there is an apple on the table, and I may desire the apple, but if my belief is mistaken I will not take the apple.  Similarly, if someone else takes the apple before I can, then I will not take it. Sometimes the effect of these factors are captured by ceteris paribus conditions which recognise that people may act differently when circumstances change, however these are generally vague and in practice circumstances will never remain constant.  We cannot expect to have law-like generalisations if we accept that they are susceptible to these other factors.

This section highlighted a number of reasons why our efforts to develop mental explanations have largely failed to provide law-like generalisations necessary for a rigorous science of the social.  However, I do not believe the endeavour is totally hopeless, and in the next section I will suggest some ways in which these difficulties may be overcome, at least to some extent.

  1. Hope for improving the power of mental explanations
In this section I will suggest three ways in which the difficulties posed by holism of the mental may be tackled, in order to allow progress within social science.   Firstly, I argue that the holism described above is not complete, and in some cases clusters of beliefs may be isolated, or dependence on particularly uncertain beliefs may be reduced.  Secondly, I suggest that evidence for certain beliefs or desires may be found via physical means, allowing objective measurement. Finally, I suggest that we may extend the scope of mental explanation to cases where other factors undermine intentional action, reducing our need for vague ceteris paribus conditions.

Holism of the mental means that every belief or desire depends for its significance on the person’s total structure of beliefs and desires.  I argue against such a complete interdependence. I accept that some beliefs and desires are fundamental, and influence many other beliefs; for example, our belief that we should generally communicate honestly, or our desire to have the things we think will please us.  However, I argue that there are clusters of desires and beliefs that do not depend on those in other clusters, and some desires and beliefs on which no others rely. In addition, I argue that some beliefs are sufficiently specific that they will seldom, if ever affect the operation of other beliefs and desires.   As a result, if we incorrectly determine some of an individual’s beliefs and desires, it will not necessarily contaminate all other beliefs and desires, or all predictions.

While the beliefs and desires within mental explanations are usually derived logically, I believe that as neuroscience develops, we will have opportunities to objectively identify some.  Already, FMRI analysis is able to determine neural correlates of aspects of our decision-making: that is, measurable physical activity in the brain that differs in a consistent way with our mental states  (Rosenberg, 2015, p68, 117). As we better understand these correlates, we will be able to make physical observations to determine what certain mental beliefs and desires are likely to be. To give an analogy from the natural sciences, the recognition that mercury expands in a consistent function of experienced heat significantly increased chemistry’s ability to accurately measure temperature and therefore experiment (Rosenberg, 2015, p67).  I acknowledge that what we currently know about the way the brain works is vastly outweighed by what we don’t know, and we most likely will not ever have full knowledge. However, it would be surprising if neuroscience’s ability to understand the basis of mental states, remained constant over the next 50 years.

Finally, rather than settling for vague ceteris paribus conditions that make mental explanations untestable in practice, I believe that considering the factors that undermine intentional action will allow social science to develop.  If we can understand specifically when lack of willpower prevents us from acting rationally, or when environmental factors will outweigh or alter our behaviour, we will be able to make falsifiable predictions and develop them into useful law-like generalisations.  For example, rather than seeing wealth maximising behaviour as something that people do just sometimes, we will be able to define a restricted set of contexts when we can genuinely expect people to act that way.

  1. Conclusion
In this essay, I described Davidson’s model of distinct physical and mental explanations for human action.  I considered how the idea of holism of the mental presents difficulties for a science of human behaviour in contrast to the natural sciences.  Although I accept these difficulties, I argued that we should not be too pessimistic, and suggested a number of ways in which I believe mental explanations can become more objective, based more on empirical evidence rather than entirely on logic and assumption.  Holism of the mental makes the mind a bit like a crossword. The task seems daunting at first, before you are confident of any letters, but solving even a few clues does make further progress easier. Human behaviour may never be as predictable as the movement of the planets, but there is good reason to expect progress.



Davidson, Donald, 2001. “Essays on Actions and Events”, Oxford University Press, Oxford

Elster, Jon, 1985. “The Nature and Scope of Rational Choice Explanation”, reprinted in Readings in the Philosophy of Social Science (1994), edited by M. Martin and L.C. McIntyre, MIT Press, pp. 311-322.

Herstein, Gary, 2005,  “Davidson on the Impossibility of Psychophysical Laws”, Synthese, Vol. 145, No. 1 (May, 2005), pp. 45-63

Jackman, Henry, 2017. "Meaning Holism", The Stanford Encyclopedia of Philosophy (Spring 2017 Edition), Edward N. Zalta (ed.), URL = <>.

Latham, Noa, 1999.  “Davidson and Kim on Psychophysical Laws”, Synthese, Vol. 188, No. 2 (1999)

Lepore, Ernest and Kirk Ludwig (eds.), 2013, A Companion to Donald Davidson, Oxford: Wiley-Blackwell.

Malpas, Jeff, 1992, “Donald Davidson and the Mirror of Meaning”, Cambridge University Press, Cambridge

Malpas, Jeff, 2015.   "Donald Davidson", The Stanford Encyclopedia of Philosophy (Fall 2015 Edition), Edward N. Zalta (ed.), URL = <>.

Pettit, P. (2002). “Three Aspects of Rational Explanation”, in Rules, Reasons, and Norms, Oxford: Oxford University Press, pp. 177-191.

Rosenberg, Alexander, 2015. “Philosophy of Social Science (fifth edition) ”, Westview Press, Colorado