20th World Congress of Philosophy Logo

Theory of Knowledge

Expertise and Rationality

Herman E. Stark
South Suburban College

bluered.gif (1041 bytes)

ABSTRACT: I explore the connection between expertise and rationality. I first make explicit the philosophically dominant view on this connection, i.e., the ‘expert-consultation’ view. This view captures the rather obvious idea that a rational way of proceeding on a matter of importance when one lacks knowledge is to consult experts. Next, I enumerate the difficulties which beset this view, locating them to some extent in the current philosophical literature on expertise and rationality. I then propose that different lessons should be drawn for rationality from the fact of expertise. One is that some empirical and phenomenological studies of the nature of expertise can be fruitfully applied by analogy to theories of the rational agent.

bluered.gif (1041 bytes)

Chicken-sexers exist. It is a fact that there are experts in this and many other domains. It is also a fact that some philosophers, often epistemologists, take the fact of expertise to be bound up somehow or other with rationality. The standard articulation of this connection is that a rational way of proceeding on a matter of importance when one lacks knowledge is to consult experts. But does this view capture the full significance of the fact of expertise for rationality? In this paper I will argue that the fact of expertise has import beyond the standard view for the analysis of rationality. I will proceed by first considering in more detail the standard view, and then jump off from that discussion to draw a different lesson for our understanding of rationality from the fact of expertise.

It would seem that the connection between expertise and rationality is so obvious as to be mundane: when one lacks expertise on a given matter of importance, and there are people who have that expertise and who can be consulted without undue cost, then (ceteris paribus) one should consult such people. This prescription links expertise and rationality by means of expert-consultation. When our cars are kaputt, when our tooth aches, when the soufflé collapses, when our French falls short, and when our business lacks efficiency we consult the appropriate experts in order to believe and behave in ways that are sufficiently non-arbitrary so as to count as rational. Expert consultation is thus a resource for rationality: we have the remarkable cognitive ability to let the outcome of expert consultation affect what we believe. (1)

The preceding paragraph captures in a simplified way the powerful idea that the connection between expertise and rationality is that of expert consultation. The idea is so persuasive that it is often assumed without reflection to be the first and last word on the matter of expertise and rationality. The obviousness of the idea is such that many a philosopher has never even considered what other connections there might be between expertise and rationality, and thus talk of expertise has been unwittingly been reduced to talk of expert consultation. Yet there are difficulties with this expert consultation idea, and there are other lessons that can be learned about rationality from the fact of expertise. Some such lessons will begin to emerge in the course of my discussion of some objections to the expert consultation idea.

Objections to Expert Consultation

A. To begin with an extreme objection, the idea that expert consultation plays a role in rationality might meet with skepticism or even derision at its mere mention. History is replete with deleterious propositions and proclamations grounded in "appeals to authority". And many a student has walked away from a logic course with the belief that any appeal to authority is an informal fallacy. Yet it is nevertheless the case that the seeking out of expertise is often part of a rational process (see Coleman 1995 for discussion of the stronger claim that there is no fallacy in arguing from authority).

It is somewhat misleading of certain logic texts to give the name "appeal to authority" to a certain informal fallacy (e.g., Hurley, 1991, p. 126). Not all appeals to authority are fallacious, despite what some undergraduates end up believing. It is perfectly legitimate, after all, for one learning to play Schumann's Träumerei to listen to a recording of Horowitz playing it. The label 'appeal to authority' should rather be reserved for a general pattern of inference which is sometimes used legitimately and sometimes used illegitimately. (2) When the correct kind of content is inserted into that pattern we might call that a case of expert consultation; when bad content is inserted we have the fallacy of argumentum ad vericundiam (see MacKenzie, 1988, pp. 57-8 for a comparison of Irving Copi and John Locke on the use of this term).

As for the history objection, one response is to note that some (e.g., John Locke in the First Treatise) have endeavored to abolish certain abused appeals to authority (often in order to allow for legitimate appeals to authority), another is to admit that even experts are fallible, and a third is to note that history is also replete with many sane propositions and proclamations grounded in appeals to authority.

B. A second objection to the consulting experts idea might be called "the wild implausibility objection". This objection holds that it would be wildly implausible to require of rational beliefs that they be actually tested against expert opinions, for many of our ordinary beliefs seem rational enough even though we do not consult experts. (3) We do, after all, manage rationally to arrange birthday parties at Sea World and change our plans upon seeing pests at the shopping mall. It is perhaps even admirable the way most of us manage to complete the dreary tasks of everyday life, given that each day we must sort through the changing circumstances to see what bits of information suggest an alteration. It seems that many of us are experts in many things — consider the difficulties Socrates would have in navigating through a modern shopping "maul". The point is that it seems ridiculous to consult experts when we ourselves are expert enough in an enormous number of everyday tasks. The point can be made more starkly by considering how we would design a being to do these tasks (see Pollock, 1989). What we would want is a being who does not need to consult others to function rationally; if the being is unable to achieve expertise in these matters, then we would not only be bad engineers, but also the creators of a pest.

The wild implausibility objection might tempt one into rephrasing the expert consultation requirement into the subjunctive mood. Instead of saying that rationality requires that beliefs be actually checked against expert opinions, one could say that such rational beliefs are those that would stand up to expert consultation. A problem with this rephrasing-suggestion is that it is not true to human fallibility. Too often we are surprised to find that what we thought would stand up to expert consultation does not.

Another and better way to meet the wild implausibility objection is simply to deny that expert consultation is a necessary condition for all cases of rationality. This suggestion waters down the expert consultation idea, but not completely. Expert consultation is one resource for being rational. As for the matter of everyday competent cognition, I will below talk of expert-imitation and suggest that this ability is one that can serve as a rational resource.

C. Another sticking point with the expert consultation idea is that in-itself it leaves mysterious the special normative status of the expert. When the layperson consults an expert the layperson assumes the epistemic authority of expertise, but what is it about the expert such that the expert's beliefs carry a special epistemic status? The answer, not surprisingly, would seem to involve training, experience, inculcating epistemic virtues, and refining perception, and I will quickly review one epistemological discussion of these matters. But what is perhaps most significant in this review is to note that accounting for the epistemic authority of the expert does not seem to entail that the expert be in possession of rules that the layperson doesn't have.

George Pappas, in his 'Experts', asks '...what is it that experts do or fail to do in their acquisition of new knowledge in their subject matters?' (Forthcoming, p. 1). Pappas examines the concept of expert as found in both ordinary usage and psychological literature. One kind of expert that Pappas identifies is the strategy-expert:

...[a strategy expert is] one who has certain strategies for acquiring new knowledge in a given subject, strategies which are different than those used by the novice and which procure speedy results (p. 5), e.g., a physicist solving a problem.

Pappas focuses on strategy-experts because he finds their abilities and achievements most intriguing, and also because empirical evidence about such experts suggests that '...we need to make room in our epistemological position for some immediately justified items' (p. 16). Pappas suggests:

Perhaps as a novice the strategy-expert went through a sequence of steps, a prescribed bit of learned reasoning that led to early knowledge acquisition in the subject area. But now, as an expert, that set of routines is no longer relied upon. Something else is operative in the case of this sort (p. 5).

Dreyfus and Dreyfus (1990) have made this same point about driving, chess, and moral development. But what is this 'something else' operating in the case of the person who no longer consults rules? A standard move is to claim that the person, now an expert, has internalized the rules; on this view an expert is a fast, fluent, and "unconscious" rule-executor. But this is not the only story that can be told; just as compatible with the evidence is the claim that in leaving rule-following the expert leaves rules altogether!

D. What ought to be done with the outcomes of expert consultation? It is sometimes rational not to accept expert recommendations, and this suggests that what is needed is a theory of what the rational relation between the expert and layperson is.

Robert Pierson (1993) claims that expertise is of two basic sorts: closed-system oriented and lay-person oriented. He uses this distinction to make the following claim: there is no rational room for lay evaluation when it comes to closed-system expert claims, but the layperson is rationally obliged to evaluate expert claims of the second sort. On the way to making this claim Pierson sets up a sort of dialectic between John Hardwig and Steve Fuller. As Pierson describes it, Hardwig holds that '...as a general policy it is more rational to defer to the authority of experts rather than to think for oneself' (p. 2). And Pierson attributes to Fuller the view that '...generally speaking, we must think for ourselves' (p. 2). Pierson ultimately argues that Hardwig and Fuller are talking past each other, and that the relationship between expert and layperson is more complex than either Hardwig or Fuller thinks (p. 20).

Pierson claims that the following three ideas are being mixed up in the Hardwig/Fuller papers:

One is the claim that we believe more than we could reasonably be expected to have the relevant evidence for, therefore we must trust others...Another is that expertise is closed system-oriented, with the implication that criticism can only be intra-systematic...And another is cost-benefit analysis, that when one decides what to do, one should act so as to maximize utility or information (p. 10).

Pierson proceeds to claim that Hardwig is mistaken in thinking that the second claim is needed to support the first claim, whereas Fuller is mistaken in thinking that Hardwig's main interest is in the second claim, and so Fuller argues against that claim by supporting the third claim (p. 10). What Pierson attempts to do is reconcile these three claims by making the distinction between closed-system expertise and layperson-oriented expertise:

When a doctor diagnoses a patient as having premature ventricular contractions, she is operating within the boundaries of her discipline — she is doing cardiology — and so there is no rational room for lay dissent. However, when the doctor recommends that the patient visit a cardiologist for care of her irregular heartbeat, she is stepping beyond her discipline of cardiology to "advise" that person about treatment. In such cases, she is no longer simply controlling and manipulating her discipline's defining set of variables, but is recommending changes to the layperson's virtually unlimited set of variables (p. 18).

In the latter case, the expert is not on a superior epistemic footing to the layperson. Pierson goes so far as to assert '...any claim with respect to how "I" ought to govern "my" life can only be rationally determined, assuming I am relatively sane, by "me" (p. 19). Although I am not willing to go this far — why can't I tell someone wiser than I my nonepistemic ends? — the main point I want to stress is that Pierson still finds a place in his scheme for expert consultation even though he admits that it is not always rational to accept the outcomes of it.

E. Here is another fairly obvious objection to the expert consultation idea: why is it rational to let expert recommendations affect what we believe when experts disagree amongst themselves? In medical matters, for example, one is often told to get a second opinion, but then almost never told what to do when that opinion contradicts the first opinion. And it is not always rational to get yet more opinions and then side with the majority of experts, e.g., when a particularly highly-regarded expert is in the minority. The point is that expert-expert interaction can have a bearing on expert-lay interaction. And thus there are two kinds of interactions at issue here. In addition to the layperson-difficulties that arise from the fact of expert disagreement, there is also the matter of what is rational for the expert to believe and do in the face of expert disagreement. For present I will offer but one remark about the layperson difficulties.

What is rational for the layperson in the face of expert disagreement? Is it rational for the layperson to conclude, upon noting expert disagreement, with the layperson's own belief? The answer here, I suggest, is most often no. The layperson often has no other rational resource available. It is rather the case that the layperson would proceed rationally by heeding an expert because either of the disagreeing expert recommendations is still superior to a random guess! (4)

* * *

The foregoing discussion should make it clear that while I am aware of objections to the expert consultation idea, I think there are lines of response — or at least partial response — available. I am therefore not out to disparage expert consultation as a resource for rationality; I think it often serves as one. But I am out to identify other ways in which the fact of expertise bears on rationality; I am not content to let the epistemic import of expertise be understood merely in terms of expert consultation.

Further Connections Between Expertise and Rationality

A first point is that the expert provides a model for thinking about the rational agent. The way in which the expert develops physical and cognitive skills, the way in which the expert arrives at conclusions, the way in which the expert transforms from the novice stage, and the flexible and adaptable way in which the expert handles novel situations are just some of the features of expertise that could and indeed should inform accounts of rationality. This general point can be illustrated by underscoring an observation that has been made by Pappas, Dreyfus and Dreyfus, Brown, and others. They note that experts in their expertise do not follow rules! However it is that experts arrive at beliefs and behaviors, it is pretty clear that they do not get there by consciously consulting algorithms. Experts are not limited to rule-following; they have a non-rule-following ability. The analogy I urge is that the rational agent should be conceived of as having some non-rule-following abilities for arriving at epistemically normative results. The rational agent, like the expert, is cognitively more diverse than a mere rule-following being. This model of the rational agent carries with it at least the following two advantages. First, the model allows us to preserve many important prima facie cases of rationality. The idea of non-rule-following rational abilities is of great help in accounting for how cognitively-limited humans manage to be rational even in those cases where rule-following is not tractably possible (see, e.g., Cherniak, 1986). Second, the model allows for a conception of an agent flexible enough to handle rationally the highly-variable and nuanced contexts of real life. Context sensitivity is just as much a part of rationality as following a set of rules.

The second lesson is that expert-imitation is just as good if not better than expert-consultation in accounting for some rational beliefs and behaviors. We do not merely consult experts, we imitate them! Kvanvig is one philosopher who stresses the importance of expert-imitation for those making their way down the path to cognitive ideality (1991, p. 172). The point I want to make is that our ability to imitate experts (and not merely to consult experts) is one resource we have for arriving at beliefs and behaviors that can be non-arbitrary enough to count as rational.

To conclude, the fact of expertise has more import than many have recognized for the theory of rationality. One general lesson for our understanding of rationality is that the fact of expertise serves to focus our attention on the range of cognitive abilities that can serve as rational resources. Experts have abilities that allow them to proceed expertly in flexible, non-rule-following ways, and this provides impetus for moving away from thinking about the rational agent solely in terms of a being that arrives at beliefs by following rational principles. The fact of expertise also provides us with a specific example of such a non-rule-following ability that can serve as a rational resource, viz., the ability to imitate experts.

bluered.gif (1041 bytes)


(1) We can, to be sure, let the outcome of idiot consultation affect what we believe, but this ability is not one which leads us to epistemically authoritative results.

(2) Hurley, to be fair, does seem to want to make a distinction between 'appeal to authority' and 'argument from authority' (1991, p. 33), but it still seems better to call the fallacy either argumentum ad vericundiam (a name which goes back to John Locke) or 'inappropriate appeal to authority'. In addition, since Hurley identifies the appeal to authority as '...a variety of the argument from authority...'(p. 126), which suggests that 'argument from authority' is the general inference pattern, it is not clear what he would call legitimate uses of the general inference pattern.

(3) Brown has been read as endorsing a strong "actual consultation" requirement (see his 1988, chs. 4-5). If this is not his view he needs to be a bit more clear there in disowning it.

(4) Professor James King pointed out to me in private conversation that the closer the layperson gets to the expertise of the expert, the less unconditional is the expert's authority over the layperson.


Brown, H. 1988. Rationality. London: Routledge.

Cherniak, C. 1986. Minimal Rationality. Cambridge: MIT.

Coleman, E. 1995. 'There is no Fallacy in Arguing from Authority'. Informal Logic, Vol. 17, no. 3.

Dreyfus, H. and Dreyfus, S. 1990. 'What is Morality? A Phenomenological Account of the Development of Ethical Expertise'. In D. Rasmussen, ed., 1990, Universalism and Communitarianism. Cambridge: MIT.

Hurley, P. 1991. A Concise Introduction to Logic, 4th ed. Belmont, CA: Wadsworth.

Kvanvig, J. 1992. The Intellectual Virtues and the Life of the Mind. Maryland: Rowman and Littlefield.

MacKenzie, J. 1988. 'Authority'. Journal of Philosophy of Education 22, pp.57-65.

Pappas, G. Forthcoming. 'Experts'. Acta Analytica.

Pierson, R. 1993. 'The Epistemic Authority of Expertise'. Preprints.

Pollock, J. 1989. How to Build a Person. Cambridge: MIT.

bluered.gif (1041 bytes)


Back to the Top

20th World Congress of Philosophy Logo

Paideia logo design by Janet L. Olson.
All Rights Reserved


Back to the WCP Homepage