Any conceptual investigation into a given phenomenon may fail in several ways. It may be, for instance, inconsistent, too inclusive or exclusive, or even materially inappropriate. In a recent reply, Michel Croce raises all of these objections to what I have called a “balanced account of expertise” (2018) [please read below the rest of the article] …
Image credit: Paco RM via Flickr / Creative Commons
Quast, Christian. 2019. “Objective Expertise and Functionalist Constraints: A Comment on Croce.” Social Epistemology Review and Reply Collective 8 (8): 15-28. https://wp.me/p1Bfg0-4jb.
The PDF of the article gives specific page numbers.
This article replies to:
- Croce, Michel. 2019. “Objective Expertise and Functionalist Constraints.” Social Epistemology Review and Reply Collective 8 (5): 25–35.
- Seidel, Markus. 2019. “Balancing the Normativity of Expertise.” Social Epistemology Review and Reply Collective 8 (7): 34–40.
Articles in this dialogue:
- Quast, Christian. 2018. “Towards a Balanced Account of Expertise.” Social Epistemology 32 (6): 397–419.
Any conceptual investigation into a given phenomenon may fail in several ways. It may be, for instance, inconsistent, too inclusive or exclusive, or even materially inappropriate. In a recent reply, Michel Croce raises all of these objections to what I have called a “balanced account of expertise” (2018). First, he claims there is a “compromising tension” between two basic components of my account (cf. sect. 3.1). This would be the charge of inconsistency, as Croce states, “Quast cannot have his cake and eat it too” (Croce 2019, 29). Second, he finds my proposal too exclusive (cf. sect. 3.2), because in his view a number of intuitive experts do not fulfill the proposed characteristics. And, third, Croce claims that two characteristics of my account should be kept apart and define different kinds of expertise rather than expertise as such (cf. sect. 3.3). This would be the charge of material inappropriateness.
The aim of this article is to answer these charges. More exactly, I attempt to demonstrate that these objections are flawed because they are based on an inadequate understanding of the framework I proposed. I will concentrate on these main objections.
2. The Framework of Expertise
A few points must first be clarified. First, in opposition to Croce’s contention, ExpertF-C-M does not represent a full definition of what it takes to be an expert, but rather states necessary conditions. This was made explicit by claiming, on the one hand, “that this framework is still no full account of expertise” (Quast 2018b, 413) and, on the other, by using “only if” formulations: “(ExpertF-C-M) Someone e is an objective expert in contrast to some client c within a certain domain d only if e is undefeatedly disposed to fulfill a particular service function in d for c adequately at the moment of assessment” (Quast 2018b, 412). It is for good reason that this framework was never intended to be a comprehensive definition; it still lacks some important features crucial for a full understanding of expertise.
Second, the framework hardly takes notice of the important fact that actual expertise ascriptions underlie two different kinds of contextuality. I could only hint at this point in the pertaining article: “Technically speaking, this can be framed within some sort of contextual indexicalism paired with the assessment sensitivity of expertise ascriptions” (Quast 2018b, 410). This neglect results from the primary aim of the article, which was to reconcile “three major dimensions of expertise ascription” (Quast 2018b, 413). It is for this reason that my emphasis was especially on the role character of expertise ascriptions, their objectivity, and defeasibility.
Third, this framework never denies that expertise is often used differently within the interdisciplinary discourse. It is indeed used in a plethora of ways. Nor does it insist that there is only a single appropriate or useful way to use the term. Instead, I strive for an explication of expertise (in Rudolf Carnap’s sense of the term) to introduce a scientifically fertile concept which largely reflects the operative usage of “expertise” within ordinary and scientific language, while setting new constraints at the same time (cf. Belnap 1993, 116). In other words, the framework suggested is a piece of conceptual engineering to promote the interdisciplinary discourse. This is why readers like Croce who prefer to continue using expertise differently are not outright wrong but are strongly encouraged to provide a similarly encompassing and fruitful alternative account.
To answer Croce’s objections, it is useful to outline four main characteristics of my balanced account of expertise.
2.1 The Objectivity of Expertise Ascriptions
Let us start with the objectivity of claims to expertise. This simply means that for an expert status to be objective the agent must be suitably disposed. This is why objective expert status is largely independent of the ascriber’s judgments and beliefs, and thus partly mind-independent. This even applies if expertise is considered an ascriptive property, that is, a property which cannot hold completely irrespective of the corresponding expert status, no matter how competent the agent really is. This objectivity of expertise ascriptions ensures that an objective expert status is grounded in the relative aptitude of a person, that is, the expert’s dispositions “to fulfill a contextually salient function adequately” (cf. Quast 2018b, 400). Thus, being an objective expert necessitates being relatively competent and relevantly inclined. To put it differently, one cannot be an objective expert if one is not suitably disposed. It is important to note, however, that this does not imply a maximalist idea of dispositions. That is, for expert status to be objectively valid the agent has to be neither Mother Theresa nor Albert Einstein. Within my framework, this requires the agent to be “undefeatedly disposed to fulfill a particular service function in d for c adequately at the moment of assessment” (Quast 2018b, 412).
As opposed to this, Croce sometimes seems to imply that expertise ascriptions are only objectively valid when involving an invariant set of competences. Although this would be desirably simple, it does not fit to the more flexible use of “expertise” in ordinary and scientific language. To illustrate, consider simple Cambridge changes. This kind of change is not based on the intrinsic properties of the object, but rather on its relational features. In this vein, the same competence of the expert of heart disease in the Zoé in South America does not constitute objective expertise at the Charité Hospital in Berlin. This is because expertise is an essentially contrastive notion. Whether one has objective expertise always depends on a specific context of use (cf. sect. 2.2). In other passages, however, Croce seems to advocate a comparative understanding of expert competence, but does not specify the required amount of relative competence. As a result, it is unclear whether he prefers an absolute threshold of relevant capacities, a simple (cf. Coady 2012, 28) or qualified (cf. Montero 2016, 59 f.) difference (to relevant contrast classes), or a combined threshold (cf. Goldman 2001, 91; Goldman 2018, 5). In contrast, I have championed a functional threshold which illuminates not only the amount of competence required, but also circumvents a series of problems for the philosophy of expertise. This is the topic of the next section.
2.2 The Indexicality of Expertise Ascriptions
Second, expertise ascriptions are often highly indexical, that is, their contents depend on the specific context of use (cu for short). Generally speaking, this context is a specific situation in which the pertaining concept is ascribed. Thus, expertise ascriptions of the form
(1) “e is an expert”
should not be understood as expressing the proposition
(2) e is an expert,
but rather as propositions of the following kind:
(3) e is an expert against a relevant client c in a particular domain d within a given context of use cu.
This is claiming that expertise ascriptions are indexical, i.e., can feature more complex and different propositional contents in different contexts of use than is immediately evident. That is, utterance of form (1) can easily have different contents. An example might illustrate this point:
Whenever I reunite with my childhood friends, they greet me with “Here’s our philosophy expert”. But this never happens to me at philosophy conferences. This can be partly explained on the basis of linguistic pragmatics, i.e., my childhood friends are mocking me. However, this should not hide the fact that they are entirely right, that in contrast to them and for the current purposes I am a philosophy expert, whereas in contrast to most of the other participants at the conference and according to more advanced standards this is usually wrong (cf. Seidel 2014, 208). In a similar vein, a general practitioner can be an expert for me in matters of cardiovascular diseases, but at the same time he or she would not be viewed as an expert in these matters by other cardiologists. This is how expertise ascriptions are bound to their particular context of use.
2.3 The Defeasibility of Expertise Ascriptions
The structure of expertise ascriptions is default and query, that is, once the status of expertise is ascribed it is vulnerable to defeat. To maintain expert status, agents need to deliver appropriately or run the risk of losing their status sooner or later. In other words, as soon as ascribed experts do not fulfill their roles, the previously impeccable expertise ascription, even if objectively valid, can be devalued or even defeated. This can happen for one of two reasons:
The expert fails either in primary or secondary matters. In the former, an ascribed expert does not execute her primary dispositions even though this would be the responsible reaction in the current situation. Consider, for example, a former football player (the expert) who for no reason refuses to comment on the match for which he was hired. In this case, it is reasonable to call his expert status into question. In the latter, in contrast, an ascribed expert does not execute her secondary dispositions properly when appropriate, e.g., an expert witness on the stand refuses to give evidence for her claims, or a member of a commission of experts refuses regular inquiries when presenting a study. Even if the expert’s testimony, or the expert’s report, is impeccable, the lacking disposition to account for their claims is prima facie good reason to defeat the existing expertise ascription.
Such a defeat can result from one-off violations of pertinent obligations, at least when they are grave enough, or can be based on repeated minor violations. In the meantime, the lacking manifestation of expertise incrementally devalues the ascribed expert status. Within this framework, an expert is therefore subject to primary and secondary obligations. But this in no way implies that experts always have to deliver in primary and secondary matters. It is rather that they have to fulfill these obligations under suitable conditions only. This means that requests to experts to responsibly fulfill their roles or queries about expert performances must also be contextually appropriate. Expert witnesses, for instance, are only obliged to serve in primary and secondary ways in court or while writing a report. Consequently, a reluctance to answer the queries of media representatives after a trial does not usually undermine their expert status. Additionally, to be contextually appropriate, a request for primary and secondary expert performances must also be empirically appropriate. As a result, experts are only obligated to do something for which they can be reasonably expected to have the required dispositions. This is because ought implies can. In conclusion, claims to primary or secondary expert performances must be both contextually and empirically appropriate.
It is this defeasibility of an expert’s status that leads to my two-stage model of expertise ascription, which can be sketched in the following way:
To be an objective expert initially requires the fulfillment of two conditions, that is, the reasoned ascription of a service function, as well as the possession of suitable dispositions for its adequate fulfillment. This not only implies competences to serve, that is, primary and secondary competences, but also a corresponding willingness to serve under appropriate circumstances” (Quast 2018b, 407).
It applies during the first stage of expertise ascription that if the expert is “unable or unwilling to give an account of her services when appropriate” (Quast 2018b, 407) her expert status may be retracted. More precisely, an ascription of objective expertise at the first stage is flawed in the former case and is gradually devalued and possibly even defeated in the latter. To put it another way, an expert must manifest her dispositions under appropriate circumstances to be considered an objective expert at the second stage.
2.4 The Assessment Sensitivity of Expertise Ascriptions
The last feature to be highlighted is the assessment sensitivity of expertise ascriptions. The content of expertise ascriptions is true or false only relative to a given standard set by the context of assessment (ca for short). Very generally, this is a situation in which a sentence is evaluated by a person. Expertise ascriptions therefore underlie contextual restrictions in two relevant ways. For one thing, they depend on their context of use. This refers to the content indexicalism set out before (cf. sect. 2.2). But for another, they also depend on their context of assessment. This is a claim about the assessment sensitivity of expertise ascriptions which is to be set out now.
The basic idea is that while the context of use is required for understanding the content of expertise ascriptions, the context of assessment is required for evaluating their truth-value. This is claiming that expertise ascriptions of the form “e is an expert in d” are true if and only if their content in cu is true in terms of the standards in ca at the time and world of cu. Summarized into a handy definition:
(ExpertF-C-M*) Someone e is an objective expert in contrast to some client c within a certain domain d at the time and world of cu against the standards of ca iff she is undefeatedly disposed to fulfill a difficult service function in d for c adequately at the moment of assessment.
That is, expertise is a complex multi-place relation, or F(e,c,d)|cu|ca for short. Assessment sensitivity leaves plenty of room for faultless disagreement in ascribing expertise, since the individual standards of assessors differ depending on their context of assessment, or, more specifically, on the expectations, practical interests, and stakes therein. This explains why expertise ascriptions are not true or false simpliciter but only relative to the standards set within the context of assessment. This also explains why there is ample disagreement when assessing expert status and why there are often ongoing discussions about expertise ascriptions in which each party claims to have and puts forward strong and even conclusive intuitions. According to my account, however, this is often faultless disagreement based on varying contexts of assessment. Rather than obsessively resolving these disagreements at a conceptual level, it might be better to illuminate their origins.
3. Croce’s Three Objections
We are now in a position to take a closer look at the objections raised by Croce. As stated, he raises an objection of inconsistency (cf. sect. 3.1), finds my proposal counterintuitive, or too exclusive (cf. sect. 3.2), and asserts that two characteristics of my account should be kept apart. This is the charge of material inappropriateness (cf. sect. 3.3). In what follows, I will demonstrate why these objections are unjustified.
3.1 The Charge of Inconsistency
According to Croce, the framework of expertise I propose has a “compromising tension” between its dispositional component and its functional interpretation. As mentioned, this framework basically understands objective expertise as the undefeated disposition to serve when appropriate, which includes not only primary and secondary competences but also the corresponding willingness to serve, that is, a relevantly virtuous character. On closer inspection, this turns out to be a fusion of dispositional elements (i.e., competences and attitudes) with a functional element (i.e., the service function). An ascription of expertise is thus objectively valid if it is grounded in these dispositions to serve and is subjectively valid otherwise. However, it is this claim which seems inconsistent to Croce, “Quast cannot have his cake and eat it too” (Croce 2019, 29). More explicitly, he claims that
it is hard to make sense of the notion of objective expertise on such a functionalist account. For possession of objective expertise in a domain becomes hostage to two inherently relative elements, namely (i) the service someone is disposed and willing to fulfill for (ii) a community—or contrast class (Croce 2019, 28).
Croce is correct in identifying these closely related aspects. For it is only on the basis of an expert’s service function that her contrast class can be determined (cf. Quast 2018b 408-11). This is hard to deny since expertise is usually understood as a relation between persons. But unlike what Croce suggests, this relationality to functions and contrast classes does not undermine the objectivity of expertise ascriptions. Rather, it makes sense to differentiate two dimensions here. Although an expert’s status is ontologically subjective within my account, since it depends on an ascription, it is still epistemologically objective; for once such an ascription has been made it depends on matters of fact whether the expert status holds or not. That is, either she possesses the required dispositions or not. Otherwise, relational properties could never hold objectively. But unquestionably, I am objectively bigger today than my father-in-law. This is simply a matter of fact and why the relational constraints under discussion do not undermine the objectivity of an expertise ascription.
However, relationality as such might not be troubling Croce but rather specific manifestations thereof. This line of argument can be found when Croce discusses my Private Expert case (cf. Quast 2018b, 410). Here, my wife asks me to find someone who can fix or replace a leaky drain pipe. To handle the issue, I ask my father-in-law, a passionate hobby-craftsperson, for support. At the end of the story, I leave my wife a message saying that “our private expert is going to solve the problem tomorrow”. The controversial question now is whether my father-in-law should be considered an expert on this matter. For Croce, he cannot be an objective expert in any case because “this diagnosis comes at the cost of giving up on the inquiry into the objective requirements of expertise” (Croce 2019, 28).
Unfortunately, he does not provide us with an argument in support of his claim, which makes it difficult to address the issue. One reason could be the widespread maximalist intuitions in using expertise “typically found in cognitive psychology. Here, experts are often considered as skilled agents who excel most others (cf. Chi 2006: 23) and are capable of performing extraordinary performances” (Quast 2018a, 16). Accordingly, experts are not only relatively-competent but also super-competent. Since my father-in-law is plausibly the former but not the latter, he should not be called an objective expert. He simply is no such super-performer. As plausible as this claim may seem initially, it faces a number of serious problems which pertain to experts of the past or the underestimated fact that we use the term in much less demanding or even minimalistic ways in ordinary language. In this vein, almost all global companies employ lawyers for international contract law who can usually be designated as the company’s experts but who are not super-competent. In view of this flexibility in using expertise, it is expedient to avoid maximalist constrictions. For expertise is a much more gradual phenomenon than Croce thinks. This is why in contrast to my wife and me my father-in-law can easily be ascribed expert status for repairing or replacing leaky drain pipes. And, as stated, this status is objectively valid iff he is undefeatedly disposed to fulfill this difficult service for us adequately at the moment of assessment. This implies neither that there are no better experts out there nor that this expertise ascription will hold from every point of assessment (cf. sect. 2.4).
In the following, Croce makes another attempt to pinpoint his objection to the objectivity of my service-functional framework. More precisely, he mentions two reasons for his dismissal:
The first is that introducing a relative element such as [(…) the disposition to serve] does not neutralize the anti-objective effect of [(…) a comparative notion of expertise]; rather, it is likely to intensify such an effect by adding a further relative variable to the account. The second is that the only way for Quast to grant expertise to his father-in-law and a plumbing engineer is to impose odd restrictions on domains of expertise […] [which lead to] an unnecessary proliferation of domains of expertise depending on the specific needs of any relevant contrast class (Croce 2019, 28 f.).
As far as the first objection is concerned, one can point to the results of our previous considerations according to which the contrastivity of the suggested framework is one matter while its objectivity is another. A conception of expertise can therefore be easily contrastive and yet allow for objective expertise ascriptions. Pace Croce, this is in no way inconsistent.
The second objection concerning the unnecessary proliferation of domains is much more promising; it is hard to deny that my framework allows for a great number of possible domains of expertise. In light of the widespread maximalist and absolutist conceptions, this certainly leads to a serious proliferation. The crucial question, however, is whether this is a vice or a virtue. To answer this question, one must not only consider this proliferation, but also the problems the framework under scrutiny is able to solve on that very basis. In my article, I have demonstrated a series of these advantages (cf. Quast 2018b, 408-11). Among the most notable is the fact that a service-functional understanding of expertise avoids the pernicious generality problem for characterizing domains of expertise that realist conceptions are notoriously confronted with: the problem of identifying valid criteria for the restriction of an expert’s domain. Without such a criterion, all else being equal, an agent can be considered an expert according to one characterization of the domain in question and yet lack this very property in innumerous other characterizations. This is due to varying degrees of reliability relative to differently characterized domains of expertise. As Croce rightly notes, though, this solution comes at the cost of a proliferation of domains. However, when taking into account that a service-functional conception of expertise not only solves this, but also a series of further problems (cf. Quast 2018b, 408-11), the complication of proliferation can be overlooked in favor of a non-arbitrary criterion to solve these problems. Thus, on closer inspection, this feature does not run against the objectivity of my framework, it rather enables it in the first place.
A negative conclusion is also obtained by looking at the first counterexample presented by Croce, in which he reports about “his auntie Renata, who helps most inhabitants of a rural village in Liguria react to (i.e., ‘like’) and comment on the content appearing in their Facebook news feed” (Croce 2019, 29). According to his understanding of my framework, Renata “would possess objective expertise in something like ‘adding likes and comments on posts on Facebook’ relative to the contrast class composed of the citizens of Bevena, although her competence regarding social networks ends pretty much there” (Croce 2019, 29). This would demonstrate that functional and objective constraints cannot be properly balanced within my framework. The framework is therefore too inclusive.
However, this judgment is premature and unsustainable for two reasons: First of all, it must be emphasized that a full conception of expertise requires the (relative) difficulty of an expert’s task. Elsewhere, I have explicated this important requirement “to avoid vexing trivializations” in ascribing expertise (cf. Quast 2018a 23). Hence, what Croce rightly points to is the intuition that having expertise always implies being “competent enough to relatively easily succeed in service-activities which are comparatively difficult for some contextually relevant reference group” (Quast 2018a 23). However, the fulfillment of this difficulty requirement is certainly questionable in the given case.
This leads us to a second rebuttal, which also rests upon Croce’s mistaken conception of my framework as representing a complete definition (cf. sect. 2). Expertise requires not only relative difficulty but also features assessment sensitive characteristics (cf. sect. 2.4). As a consequence, expertise ascriptions like “Renata is, relative to the citizens of Bevena, an expert for adding likes and comments on Facebook posts” are only false against the relevant standards of evaluation within the present context of assessment. For this reason, it is faultlessly possible that, based on two different contexts of assessment, such an expertise ascription is true as assessed from one context and perspective and false from others. For example, imagine that one assessor sets a higher standard for difficulty because he follows maximalist intuitions, and he primarily seeks scientific expertise in the search for the best possible expert commission, whereas another assessor favors lower standards because she seeks expertise as expressed in ordinary language, where not much is at stake. This does not mean we are losing grip on an objective notion of expertise, but simply reflects the fact that the use of “expertise” is much more flexible than Croce would like to admit.
Thus, objective and functional constraints fit properly together. This is not even a big surprise. For, on closer inspection, the relational features of my framework that Croce objects to do not undermine its objectivity, but only run against an absolutist understanding of expertise. This merely implies the largely uncontroversial claim that there is no single, intrinsic, and invariant property identifiable in expertise.
3.2 The Charge of Overexclusiveness
A framework can fall victim to several defects. It can include cases that should be intuitively excluded, i.e., the charge of overinclusiveness, and it can exclude cases that should be intuitively included, i.e., the charge of overexclusiveness. The former has already been discussed, so the latter charge will be examined now. Croce claims that my framework imposes conditions that make it impossible for many individuals who would be intuitively ascribed expertise to fulfill. “In particular, I worry about secondary competence and what Quast calls ‘intellectually virtuous character’ as necessary components of expertise” (Croce 2019, 29). It was stated earlier that secondary competences are those dispositions which apply to the service-functional exercise of primary competences (cf. Quast 2018b, 414, fn. 21). Croce tries to illustrate this concern by means of two different counterexamples:
In the first case, he considers a highly skilled civil engineer with the secondary competence “to discuss [(…) his actions] with other experts yet lacks the competence to provide effective explanations of [(…) relevant] techniques, strategies, and related risks to a lay audience” (Croce 2019, 30, my italics). Croce considers the civil engineer an expert, whereas in my framework this classification would be impossible. However, this is highly questionable for several reasons.
Firstly, I did not argue against the common intuition that a lack of primary competences is often much more damaging to an expert’s status than a lack of secondary competences or the willingness to give an account of one’s performances (cf. Quast 2018b, 416, fn. 41). This is because a lack of primary competences is often much harder to eradicate than a corresponding lack of secondary competences. Thus, “primary” is not only a predicate of order, but also evaluative to a certain extent.
Secondly, my demand for secondary competences should not be misconceived in maximalist terms. Rather, experts should be able to explain, justify, and respond to their performances, at least in general terms. But this neither implies that they can perfectly expose every single detail of their capacities to every possible contrast class nor that they consciously understand every single detail of the outcome. Instead, the required degree of secondary competence is always a function of what can be rationally expected, the service ascribed, and the applied standards of assessment. A more precise determination of the required degree of secondary competences thus requires greater power of judgment on behalf of the assessor. To require more comprehensive or more precise specifics here would lead to an over-intellectualization of expertise. Rather, my claim for secondary competences reflects the common intuition of experts passed down through centuries, “If some practice is an expertise, then it can be taught and learned, and any true expert should be able to teach his expertise” (LaBarge 1997, 53).
Thirdly, I just argued that if an alleged expert is not able or willing to explain his course of action when this would be the responsible reaction, there is good reason to retract a prior expertise ascription due to defeat. But whether a devaluation or retraction takes place is ultimately a decision of the assessor of the expertise ascription. And because different standards of accountability can be imposed on experts, faultless disagreement can occur. This is claiming that the shown accountability of an expert is an assessment-sensitive matter of fact which could be judged differently. It is for this reason that Croce’s counterexample of a skilled civil engineer is unsuitable for undermining the framework under scrutiny.
In the second counterexample, Croce refers to athletes who “can do extremely complicated things, yet they may not be able properly to account for what they do” (Croce 2019, 30). Indeed, this is a well-confirmed empirical finding: “[E]xperts often cannot articulate their knowledge because much of their knowledge is tacit and their overt intuitions can be flawed” (Chi 2006, 24).
Let’s consider an example. When experts are asked to explain their performances, these explanations are often inconsistent with their observed behavior, that is, the application of the given rules does not lead to similar results in simulations. To a substantial extent, this circumstance can be traced back to the relative independence of two kinds of cognitive processing (cf. Stichter 2015, 111-3). On the one hand, there are the automatic, intuitive, effortless, and quick operations of System 1 on which elite performance is usually based. On the other, there are the conscious, analytic, effortful, and slow operations of System 2, which are employed while explaining one’s performance, for instance (cf. Kahneman 2014). Thus, cognitively considered, to perform well is one thing, to explain well is another. It is therefore to be expected that elite athletes will not always be equally good experts in their field. As opposed to this, by putting forward this second counterexample, Croce suggests that elite athletes are always experts for the respective domain of action.
Against the background of the framework under scrutiny, one could respond to this claim in two different ways. One could simply insist that there is indeed a crucial difference between elite athletes on the one hand and experts on the other. The former might be highly competent, but they are not experts because of their lack of relevant dispositions to serve. However, it might be a better idea to balance the opposing intuitions here. To achieve this, we need to take a deeper look at the framework in question. More exactly, it claims that an expert needs “to elucidate her performances when appropriate in order to retain her competence and responsibility” (Quast 2018b, 406). We must now focus our attention on the important “when appropriate” adjunct. Although explicating this aspect was not the focus of the article in question, it might have avoided some unnecessary misunderstanding. So, it is Croce’s critique which welcomes the opportunity to add some clarifying remarks here (cf. sect. 2.3).
On closer inspection, this adjunct ensures that queries about experts and expert performances are fundamentally and contextually appropriate. For one thing, this means that members of contrast classes are not automatically entitled to call for an explanation, justification, etc. Rather, the query itself must be contextually appropriate. To exemplify this, recall that an expert witness is obliged to serve in court only in a secondary way; she is not, say, obligated to respond to media representatives after the trial. However, queries need to be more fundamentally appropriate as well. This posits that we expect experts to possess the required dispositions in the first place. Yet, this requires that these secondary dispositions are the rule rather than the exception, that is, are empirically expectable at least. As we have already seen, however, it is a well-established finding that not only elite athletes have problems giving detailed accounts of their performances. This should not lead one to the delusive conclusion that this shortage of pronounced secondary competences dispenses elite athletes from explaining crucial features of their performances to relevant contrast classes when appropriate. Rather, if they are objective experts, they need to be responsive to this requirement, at least to a certain extent. This is no excessive claim. For elite athletes usually do have a much better understanding of their own performance and the fundamentals of the sport than a relevant laity, for instance. It is thus crucial to note that the requirement for secondary competences is always a matter of degree and must be both empirically and contextually adequate. It is for this reason that the second counterexample also fails to undermine the framework of expertise discussed here.
3.3 Keep Apart What Needs to be Distinguished
So far we have seen that a balanced account can be easily defended against the aforementioned objections. As it turned out, it is consistent and neither overtly too inclusive nor too exclusive. In view of these results, it’s time to take a closer look at what drives Croce’s critical stance. Which intuition might underpin his reasoning? As I take it, the fundamental bottom-line is his claim that “primary competence and secondary competence are, in a sense, different kinds of expertise” (Croce 2019, 30). Such a view differentiates between “the expertise of those who can reliably provide some sort of service in a domain and those who can explain what’s going on in a domain to others, especially laypeople. Call the former domain-oriented expertise and the latter novice-oriented expertise” (Croce 2019, 33). This distinction is certainly useful for some heuristic purposes. It follows a series of very similar distinctions that have already been made elsewhere:
i. intellectual expertise and practical expertise (cf. Goldman 2001),
ii. adaptive expertise and routine expertise (cf. Hatano and Inagaki 1984) and especially
iii. interactional expertise and contributory expertise (cf. Collins and Evans 2007).
Without going into detail here, these distinctions can be attributed to a basic difference in human cognition, namely, Kahneman’s (2014) distinction between the automatic, intuitive, effortless, and quick operations of System 1 and the conscious, analytic, effortful, and slow operations of System 2 (see section 3.2). This means that the intellectual, adaptive, interactional, or novice-oriented expert primarily uses System 2 operations, whereas the practical, routine, contributory, or domain-oriented expert relies on System 1 operations.
These are, unquestionably, revealing differences grounded in the specific nature of information processing; they are particularly useful for understanding observable features of expertise especially in the disciplines of differential psychology, epistemology, and sociology. However, the question arises whether these differences are also conceptually fundamental. This is where my position diverges from Croce’s. According to him, objective expertise requires either primary or secondary competences, while my position requires the simultaneous existence of both, which is informed by the actual use of expertise in ordinary and scientific language. More particularly, my position rests upon a balancing of three different dimensions of expertise ascription, that is, the role-functional, dispositional, and manifestational dimensions. What is most relevant here, however, is the close interrelation between the role-functional dimension on the one hand and the dispositional one on the other. This has led to the thesis that expert competences are best understood as competences to serve (cf. Quast 2018b, 408 ff.). As explained previously, this not only requires primary competences which refer “to technical skills, knowledge or understanding of a contextually relevant subject matter (plumbing, particle physics, epistemology, wine tasting, etc.)” (Quast 2018b, 414, fn. 21), but also secondary competences applying “to the social exercise of primary competence[s]” (Quast 2018b, 414, fn. 21). Here, for example, one might think of “performative abilities, mediation competences, communication skills, trust-building abilities, etc.” (Quast 2018b, 414, fn. 21). Secondary competences are viewed as those capacities which put primary competences to work in the first place in social environments; put another way, they are capacities “to establish and retain mutual trust between experts and clients” (Quast 2018b, 407). It was also demonstrated there that such an understanding is not singular but implicit within a number of other characterizations of expertise (cf. Quast 2018b, 414, fn. 21).
For the sake of simplicity, my article focused on the accountability of experts, which was also due to the epistemological background of the publishing journal. However, the article also mentioned that experts “must be disposed to properly estimate and communicate the scope and limitations of their competences” (Quast 2018b, 407). That these requirements are in place becomes obvious when experts repeatedly or grossly violate their obligations to serve. To illustrate, recall that my father-in-law can easily lose his expert status while staying relatively competent nevertheless. This can happen, for example, if he repeatedly refuses to give an account of his approach when appropriate, or if he does not responsibly estimate and communicate the scope and relevant limitations of his competences to my wife and me. Moreover, consider a pianist who is able to play Mozart’s Sonata No. 18 in D major flawlessly and with great devotion, but suffers from pathological stage fright and therefore avoids public performances. According to the framework under discussion, such a person cannot be an orchestra expert on piano playing, since she does not have the required secondary or performative abilities. As a consequence, it is also imaginable that highly advanced computer security geeks who are completely detached from ordinary communication lack objective expertise for you and me, but can still be considered objective experts for concerns dealing with computer security at the specialty department of Alphabet Inc., for example. The claim is not that these geeks are less competent in primary matters. My suggestion is just that they should not be considered objective experts for you and me, for they do not possess the required competences to serve. If this is convincing, then it is strong evidence for the claim that expertise requires secondary competences, whether these are explanatory, performative, or of another nature. Those who do not responsibly manifest a secondary competence under appropriate circumstances will sooner or later lose their expert status. This is a reference to the third manifestational dimension of expertise ascription mentioned above.
Croce’s objections prove to be unfounded. They fail due to an inadequate understanding of the underlying framework and its implications. More precisely, it was shown that this framework allows for expertise ascriptions which are objectively valid (cf. sect. 3.1), can handle a series of proposed counterexamples easily (cf. sect. 3.1-2), and can explain why it is suitable to consider expertise a disposition to serve, comprising not only primary, but also secondary competences and attitudes (cf. sect. 3.3). However, to avoid any unnecessary misunderstandings, it should be repeated that the controversial requirement for secondary competences was found to be subject to gradual, empirical, and contextual restrictions. Thus, the specific kind and extent of requirement always depends on the plausible assignment of a service function, that is, on its context of use, and even partly on the given perspective of its assessor within the salient context of assessment. Ironically, it is this assessment sensitivity of expertise ascriptions which allows for faultless disagreement within different contexts of assessment. And this implies, to the chagrin of Croce, that diverging intuitions regarding my Private Expert case can be well justified (cf. Croce 2019, 30).
To sum up, it now makes sense to distinguish two crucial features. First, there is the adequacy of a given ascription of expertise and, second, the adequacy of a given definition. The point of this distinction is that two assessors of an expertise ascription may easily diverge on the former yet agree on the latter. The bottom line is that Croce’s counterexamples fail to undermine the proposed framework and may even exemplify it. The underlying disagreements are even so widespread that examining it on a conceptual level is not a vice but rather a virtue.
Contact details: Christian Quast, Westphalian Wilhelms-University of Münster, firstname.lastname@example.org
Belnap, Nuel. 1993. “On Rigorous Deﬁnitions.” Philosophical Studies 72 (2/3): 115–146.
Chi, Michelene T. H. 2006. “Two Approaches to the Study of Experts’ Characteristics.” In The Cambridge Handbook of Expertise and Expert Performance, edited by K. Anders Ericsson, Neil Charness, Paul J. Feltovich, and Robert R. Hoffman, 21–30. Cambridge: Cambridge University Press.
Coady, David. 2012. What to Believe Now: Applying Epistemology to Contemporary Issues. Chichester: Wiley-Blackwell.
Collins, Harry M., and Robert Evans. 2007. Rethinking Expertise. Chicago: University of Chicago Press.
Croce, Michel. 2019. “Objective Expertise and Functionalist Constraints.” Social Epistemology Review and Reply Collective 8 (5): 25–35.
Geach, Peter T. 1969. God and the Soul. London: Routledge and Kegan Paul.
Goldman, Alvin I. 2001. “Experts: Which Ones Should You Trust?.” Philosophy and Phenomenological Research 63 (1): 85–110.
Goldman, Alvin I. 2018. “Expertise.” Topoi 37 (1). Springer Netherlands: 3–10.
Hatano, Giyoo, and Kayoko Inagaki. 1984. “Two Courses of Expertise.” Research and Clinical Center for Child Development Annual Report 6: 27–36.
Kahneman, Daniel. 2014. Thinking, Fast and Slow. New York: Farrar, Straus and Giroux.
LaBarge, Scott. 1997. “Socrates and the Recognition of Experts.” Apeiron 30 (4): 51–62.
Merton, Robert K. 1942. “A Note on Science and Democracy.” Journal of Legal and Political Sociology 1 (1-2): 115–26.
Montero, Barbara G. 2016. Thought in Action. Oxford: Oxford University Press.
Quast, Christian. 2018a. “Expertise: a Practical Explication.” Topoi 37 (1). Springer Netherlands: 11–27.
Quast, Christian. 2018b. “Towards a Balanced Account of Expertise.” Social Epistemology 32 (6): 397–419.
Seidel, Markus. 2014. “Throwing the Baby Out with the Water: From Reasonably Scrutinizing Authorities to Rampant Skepticism About Expertise.” Informal Logic 34 (2): 192–218.
Seidel, Markus. 2019. “Balancing the Normativity of Expertise.” Social Epistemology Review and Reply Collective 8 (7): 34–40.
Stichter, Matt. 2015. “Philosophical and Psychological Accounts of Expertise and Experts.” Humana.Mente Journal of Philosophical Studies 28: 105–28.
 A word in quotation marks (“expertise”) designates, as usual, the word as a word. If not otherwise contextually apparent (by phrases like “the notion/explication/ascription of expertise”) an italicized and bolded term (expertise) refers to the concept expressed by that word, while regular print (expertise) designates the respective phenomenon.
 The term is coined by Peter Geach (1969, 71 f.). A simple example of such a change is the following: Imagine that Tom has a hip measurement of 116 cm, while Jim’s is only 82 cm. After a couple of years, Tom’s hip measurement stays the same but Jim’s increases to 119 cm. As a result, Tom initially has a larger hip size than Jim, but then loses this property without a change in his intrinsic properties.
 Generally, the context of use is a specific situation in which the pertaining concept is ascribed.
 I have repeatedly argued for this, cf. Quast (2018a, 2018b).
 See also Seidel’s (2019, 35 f.) instructive comparison of two kinds of normativity within my framework (i.e. the dispositional and the functional) and Merton’s similar approach to the normativity of science (cf. Merton 1942).
 So, my father-in-law need not account for the same level of detail like a professional plumber. Rather, it is the case that normally greater expertise also implies greater obligations.
Categories: Critical Replies