by David Foss
To make a claim or assert as fact something or other, is to commit oneself to the defense of that claim or fact. To claim something as true is to be committed to the justificatory task with regard to that thing upon objection. The substance of such a defense need not be terribly complex, and may even involve (indeed, frequently does involve) a high degree of deferential hand waving. That is, when I defend a claim like “a molecule of water is composed of one oxygen atom and two hydrogen atoms” (expressed simply as “water is H2O”), I am committed to the justificatory task upon objection, even though a great deal of that task (for me) will involve statements like “Dr. Linus Pauling, a reliable Chemist, has informed me that this is the case.” The fact that I defer a portion of the justificatory burden to individuals I trust does not alter the fact that it was my burden to defer. If the degree of deference is high enough, we may wish to deny that the speaker has “knowledge”; but it would be difficult to find a case in which at least some degree of deference is not present in a substantial fact claim. By deferring to the position, fact claims, experimental reliability, ontological categories, etc., of another scientist, however, I am not ambivalent to the truth of their claims (at least those upon which the legitimacy of my fact claim rests). By claiming something as true, and by passing some of the justificatory burden on to others, I am endorsing that portion of their work which supports my claim.
In normal discourse, this “passing on” of the burden often does not single out the work of individual scientists. It frequently involves an appeal to “accepted scientific wisdom.” And here the situation is not terribly different from that where a scientist or theorist is mentioned by name. Belief in the reality of atomic and sub-atomic particles, whose general behavior is relatively well known and documented, represents a broad field of knowledge capable of supporting all sorts of claims about molecular physics, even where the speaker is unable to name the scientists whose work helped legitimize the principles upon which her claim rests. But here, as well as in those cases where the speaker can name the persons to whom she defers, the reliance upon a body of work, to carry out a portion of the justificatory task, essentially involves an endorsement of that work, and those aspects of that work necessary for the establishment of its legitimacy. When I claim that “water is H2O”, I cannot avoid endorsing the basic picture of the micro-cosmos held by Niels Bohr near the end of his life, even if I do not know his name. Even more importantly, I cannot avoid endorsing (without some rather extravagant dialectical gymnastics) Bohr’s basic instrumental and methodological assumptions: the assumptions about what his instruments are measuring, and what sorts of objects (in the world) could effect the outcome of his measurements; and the assumptions about how we ought to go about examining the world, and who we are to engage in such an examination (i.e., how it is that we can and ought to pursue the world in such a fashion). I am, wittingly or not, endorsing all this in my absurdly mundane fact claim, “water is H2O.”
In more straightforward justifications of fact claims, where deference is not explicitly acknowledged (eg., any case in which I do not appeal to scientific dogma, or the views of a specific scientist, as such), there remains a substantial degree of appeal to methodological and ontological assumptions whose legitimacy resides in common scientific practice. The integrity of claims which refer to “air”, “sound”, “time”, “space”, etc., rely upon basic beliefs about the constitution of the world which have been fundamentally changed by developments in aeronautics, acoustics, the theory of special and general relativity, and a plethora of scientific sub-disciplines over the last one hundred years. Claims concerning the condition and status of ordinary objects, from chairs and pets to touches and illusions, have all been effected (and therewith infected) by the modern scientific way of carving up the world. Wittingly or not, we rest our fact claims on a world made, in large part, by science. Still, we might wonder about the degree to which a justificatory commitment is a fundamental ingredient in non-scientific fact claims.
The assertion that any fact claim will necessary involve a commitment to the justificatory task, may sound a bit over-stated. Some fact claims, even some fact claims about the world we live in, seem so obvious that they are in no need of being justified by other methodological and ontological commitments (commitments to the right-ness of particular truth-testing methods and the reality of specific sets of world-objects). I am warranted in holding as true the views that “snow is white”, or “sugar tastes sweet”, in spite of the fact that I may be unable to articulate a theory as to why snow appears white, or sugar tastes sweet. The only world-objects named, and the only ones I need to name, are complex objects: snow and sugar do not (or need not) name ontologically simple entities, and so are ambiguous between a wide array of ontologies. They are entities in the Sellarsian “Manifest Image,” and may reliably speak from an almost limitless variety of “Scientific Images”. The qualities of “being white” or “tasting sweet” are likewise ambiguous between widely divergent world- views. And most of the differences, between ontologies in which snow and sugar exist and between world-views in which the predications of being white and tasting sweet make sense, make no difference as to whether we count as justified the view either that “snow is white” or “sugar tastes sweet”. Indeed, we will more likely count as mistaken a scientific view which is unable to justify “snow is white”, than count the claim “snow is white” as false if contradicted by a particular scientific interpretation.
Does any of this show that there exist fact claims which do not call for a justificatory commitment? I think not. Strategies of justification may shift substantially as we move from the context of “common sense” assertions to a context of complex scientific assertions; but the justificatory burden is as present in the prior as it is in the latter. Witness a conversation in which the commitment is abandoned:
Henry: “There’s a hole in the bucket.”
Leisa: “No there isn’t.”
Henry: “Hmm. But there is.”
Leisa: “I don’t see a hole.”
Henry: “But there’s a hole in the bucket.”
Leisa: “Oh, you’re just saying that.”
If I refuse to defend my claim, or just state that “it’s true” without offering reasons, I will rightly be recused for “just saying that”. So long as Henry avoids showing Leisa the bucket (certainly a form of argument in this context), or showing her things the bucket cannot do because of a hole (like hold water), he has not committed himself to the truth of the claim “there’s a hole in the bucket.” Until he commits himself to the defense of his view, he will not be claiming it as a truth. And until this occurs, the claim cannot even appear as a candidate piece of knowledge.
The critical thing to notice, however, is not so much the presence of a justificatory burden, as it is the presence of a tacitly endorsed justificatory strategy. In any fact claim, accompanying the speaker’s submission to the justificatory burden, there resides a substantial view about how we ought to go about defending our views. In any fact claim there resides an endorsement of a particular knowledge-making process, appropriate to the context in which the claim is made. Not incidentally, the endorsement of such a process essentially involves a view about who we are that such a process is both possible and good.
It is critically important to recognize that the constitution, and even the presence, of such methodological endorsements are frequently submerged beneath common sense notions of “rightness” and “truth”. Henry will likely have no idea that he is committing himself to a particular sort of justificatory strategy (as being good for us) when he shows his bucket to Leisa. But he would again fail to be making a fact claim if, upon questioning, he revealed that he “couldn’t care less” about the legitimacy of “seeing” as a reason for believing there is a hole in his bucket. Likewise, if upon questioning, he admitted that he believed that people could not engage in such visual proofs, or that such proofs were not good to engage in, we would be warranted in withdrawing our judgement that he was making an honest fact claim. He would merely be “playing with us,” or worse.
Common sense fact claims are interesting, but they tend to take for granted the methodological and ontological commitments which make them thorough-going (“honest”) fact claims. In general, Henry will not notice his strategic endorsements, and we will not draw his attention to them, during the interrogation of particular claims about the world. But if Henry is a chemist (or a physicist, or biologist, etc.), the matter of methodological and ontological commitments becomes a more obvious (and reflectively self-conscious) part of his assertions. Whereas Henry of the Bucket may be said to endorse a particular justificatory method simply by his use of it, Henry the Chemist will also be expected to be able to tell us about his method, and why it is the appropriate one to use. Henry the Chemist will not only be aware of the fact that he is endorsing a specific justificatory method. He will be able to tell us (and not merely show us) what this method consists in. At least, this is what we may expect of a practitioner in the discipline of a modern science.
In the present context, the reasons for this are not as important as the fact that it occurs. Scientists take part in a discipline which seeks to articulate, and not merely practice, the heuristics of knowledge. The sciences, from physics to paleontology, are not alone in possessing such a double duty. Still, they are generally considered the disciplines which have most succeeded in both respects. I speak of science because we are speaking of fact claims and knowledge. And for us, to a large extent, science is the dominant practice of knowledge generation. The hope is that we can generalize the results of the current analysis to other “knowledge-realms”, and other “knowledge” related disciplines, such as philosophy, history, politics, and so on. If it is true that fact claims in clinical contexts are forever tied to essentially normative claims about who we are, then truth-claims in even more humanistic contexts will surely possess at least that degree of normativity.
But is this a sort of commitment, with respect to specific methodological and ontological assumptions, which essentially involves an irreducibly normative take on who we are? Might the methodological commitments I undertake by the utterance of a fact claim be entirely understood without any mention of who I am (or how I ought to be treated)? There are good reasons for wanting to make claims about the world which do not rely upon fundamentally normative claims about the constitution of knowers. It would be nice to have some body of knowledge about the world which cannot be accused of depending upon some rather wishful thinking about our epistemic capabilities or our role in the universe. But there are also some not so good reasons for resisting the introduction of deep valuational commitments into the content of fact claims.
One might worry that given such bedrock normativity, with respect to the constitution of knowers (participants in the scientific project), it will be impossible for a science (or any knowledge- discipline) to consistently hold that any conclusion about the constitution of persons is non-normative or “objective”, such that a primitively normative challenge is necessarily invalid. Indeed, it would appear that no position with regard to our world, no matter the volume of statistical evidence which supports it, could avoid all primitively normative challenges. Even the most “objective” of our scientific claims would rest, at least in part, on some premise for which no non-normative defense could be given. Such foundations would forever open even the most trivial truth-claim to an essentially political objection. The more tempered of these fears, concerning our capacity (or our substantive inability) to make non-normative claims about persons, seems to me correct (and a bit apropos). The more extreme, concerning the degree to which such an analysis needlessly burdens scientific work with the spectre of objections from “social realism”, is importantly over-stated.
We may admit that our capacity to make sensible claims about the world rests on at least some deep assumptions about how we ought to behave, and that the “ought” of such assumptions cannot be turned into the “must” of natural necessity, without claiming that empirical matters pale in the shadow of such normatively charged underpinnings. Here we need only recognize that an investigation into the normative aspects of fact claims need not diminish the degree to which we hold that, most of the time, empirical matters really ought to dominate factual disputes.
There is some extent to which we really should view with an abiding skepticism any science which attempts to tell us (too explicitly) how we ought to think of ourselves and our epistemic capabilities. Nevertheless, if the current analysis is correct, any science, as a knowledge-practice, will involve some attempt to tell us just this. It may be an unfortunate tendency of highly sophisticated sciences to offer their view of persons (where they offer one at all) as a result, and not as a premise, of their practice. The degree of sophistication involved in making veridical claims in such practices may help cloak the degree to which claims about persons (at any level) remain essentially bound to the practice’s normative (i.e., political) base. A normatively charged view of persons will reside in the deepest presuppositions of any knowledge practice. Most of the time, we can safely ignore this fact, concentrating our attention instead on the mechanics of the system. But when the system produces a result (or is said to produce a result) which concerns the proper treatment of persons, or which prescribes some behavior to us (as persons), we cannot afford to rule out a critical review of the system’s prior political commitments.
But what shall the commitments amount to? So far, a great deal has been said about the existence of such commitments. Very little has been said about the necessary (or natural) scope or content of such normative presuppositions. Let us retrace our steps briefly, and extend the analysis.
When I state as fact some thing or other, I am committing myself to the legitimacy of my claim. This legitimacy resides in a strategy of justification which I do or would pursue if my fact claim were challenged. I cannot be indifferent to the legitimacy of this strategy. Indeed, by its actual or expected employment, I must be committed to the defense of its legitimacy as a strategy for procuring the legitimacy of my initial fact claim. Now, the legitimacy of my justificatory strategy must concern two things. Firstly, my defense must demonstrate (or successfully argue) that this is a strategy which seeks to obtain for us a thing we are interested in having. Secondly, it must demonstrate that this is a strategy we are capable of employing successfully. The two terms to notice here are: in the first case, the attention to our interest as knowers; and in the second, the matter of the successful employment of my justificatory strategies. Both of these seem to me inextricably normative in content. That is, there seems no way to express what our relevant interests are, or what the successful employment of justificatory strategies will consist in, without appealing to some prior notion of social propriety: the way people ought to act with respect to their world and each other. It may be hard to see immediately how the second of these commitments (concerning the success of a truth-project) depends upon irreducibly normative claims about the good life. Fortunately, the trail can be blazed reliably by an investigation into the first (concerning our interests in practicing a particular knowledge-craft), where the link between “interest” and bald normativity is more accessible.
Let us turn to the matter of interest. Here we are asking what it is that I affirm about my interests in playing a particular justificatory game: the game of asserting, defending, and criticizing scientific or fact claims. What must I be committed to in order to satisfactorily claim that this is a strategy which seeks to obtain for us a thing we are interested in having. At the very least, we might say that I am committing myself to the relatively simple view that I, as a scientist (among like-minded scientist-persons), am a truth-seeker. Whatever else there is to the human psyche, the part which matters to my fact claim (and therefore needs to be an ingredient of the psychological view I concomitantly endorse) is the capacity and desire to learn about the world.
At another extreme, it may be that I am committing myself to a wide array of historical assumptions about persons, some of which I would find offensive if brought to my attention, but all of which I must endorse in order to consistently believe in the legitimacy of my justificatory strategy. When I take a stand with respect to the legitimacy of a practice, by the very act of engaging in that practice as grounds for the legitimacy of what I have to say, I embrace a logic (and world-view) the ramifications of which I may only vaguely comprehend. There is no a priori reason that these ramifications should be limited to those aspects of the human character which directly concern our status as knowers, even though the strategies we embrace on our way to truth only speak explicitly of that part of us which knows.
But let us turn our attention first, to the simpler possibility. It would certainly simplify matters if the only sort of political or primitively normative claim necessary for the legitimation of fact claims was any one which endorsed our role as truth-seekers. It would then be the truth-seeker-ness of our psychological presuppositions which bore the principle justificatory weight in defending our interest in particular “knowledge” procuring strategies. To test the legitimacy of a specific strategy, all we need determine is whether it relies upon a conception of persons which reserves a central place for the acquisition and processing of truths. The remainder of one’s conception of personhood is only relevant to the degree that it infects the more directly truth-related faculties. So long as one’s picture of personhood can consistently view people as capable and inclined to seek the truth, endorsement of the package will provide a sustainable defense of routine knowledge claims.
Unfortunately, how we seek truth is not just about truth. It is about a way of life. It is about living “in the light of truth”; and living “with a mind toward truth”. As a way of life, the seeking of truth cannot be non-partisan with respect to various interpretations of personhood and the moral status of participants, observers, and even “outsiders”. More importantly, however, the truth which we are herewith said to seek no longer appears capable of constituting even a portion of the shared psychological presuppositions needed to get the knowledge game going. In other words, viewing truth in terms of a way of life (essential to the relevant interest of “seeking truth”), requires that we approach the matter of truth in a way which does not allow us to see “truth” as identifying a specific sort of human interest. And this will mean that “seeking truth” does not identify a (distinct) category of psychological theory which must be endorsed in ordinary fact claims. But why does truth collapse when viewed through the parallel lenses of ‘interest’ and ‘a way of life’?
Our context here is that of the act and process of justification. Whatever virtues there may be to viewing “truth” as a bona fide property in other contexts, there seems none relevant to the phenomenon of making truth claims. Tarski’s disquotational theory of truth is nowhere more at home, or more sensible, than in the discursive process of justifying and explaining claims about the world.
Admittedly, nouveau-pragmatists, like Sellars, Brandom, Camp, etc., do not think a notion as complex as truth is exhaustively captured by Tarski’s schema. Still, truth itself plays little to no explanatory or justificatory role in our common scientific practice, just as it plays only a nominal role in the explanation of generic fact claims. The claim ‘‘p’ is true’ is not a robust defense of the claim that ‘p’. Indeed, it hardly seems to be a defense at all, but merely a retelling of the same assertion (in slightly more direct or forceful terms — as if to make the claim a more emphatic one). In almost all circumstances in which fact claims are asserted and defended, no insight is gained by the insertion of the words “is true”. Other strategies carry the weight of actual explanation and justification. It has often been tempting to equate these strategies with the meaning of “truth”, such that the “is true” predicate would stand for (or signify) the successful execution of current justificatory practices. Expressed this loosely, it would seem that there would be considerable merit in such a move.
Still we might ask, what is there to “is true” that is not captured in primitive assertions? On some occasions, by including the predicate “is true”, there seems to be a sense that the justificatory project has to a significant extent been completed. By merely asserting p, I am committed to defending, upon challenge, the reasons for believing p. It may be the case that I have not worked out the details of the requisite justifications, without thereby undermining the validity of my present commitment to p. But by asserting ‘p is true’, I seem to be making the stronger claim that I already have good reasons for supposing that p; it suggests that I (or persons I am entitled to trust) have already done most of the justificatory work, or that I simply haven’t much justificatory work left to do.
This distinction, signaled by the presence of the “is true” predicate, may be captured by the admission that asserting ‘p is true’ is simply a more emphatic way of asserting ‘p’. However, the existence of such a distinction suggests that “is true” cannot be eliminated from fact claims, without altering the meaning of those claims, even if the difference does not directly impact upon our sense of what there is (i.e., reality). Importantly, the nature of the distinction also suggests, that the “is true” predicate does not attach particular justificatory strategies to a claim. It only signals confidence that such strategies have been explored and satisfied. The simple assertion ‘that p’ already contains a commitment to justification, and contemporary strategies of justification are therewith (implicitly) attached to the assertion.
What does all this tell us about the notion of Truth itself? Is truth the sort of animal we can seek by our justificatory strategies? Of course, there are many things we might mean by the title Truth. Most of the candidate meanings of such an idea are clearly not the sort of thing which can be intelligibly sought. If by truth we merely mean reality (according to most substantive forms of realism), then should we say that it is the search for the real world that justifies our verificational or justificatory strategies? Perhaps we seek an understanding of reality, but can we “seek” reality itself? What would such a search even look like? (It would have to be a reality we do not already inhabit or constitute, and about which we could only have indirect knowledge. Plato might be pleased by such a picture, but “reality” now becomes as fuzzy a notion as truth.) If we look back to our brief exploration of the predicate ‘is true’, we might suspect that “Truth” stands for the class of propositions I am entitled to affirm, or (at the very least) I have good reasons to believe. However, looking more carefully at the function of the predicate sketched above, “Truth” could at most only stand for the acknowledgement that, with respect to a given set of assertions or claims, contemporary strategies of verification and justification have been satisfactorily applied. This no longer seems an attractive goal for the scientific endeavor. More importantly, however, this Truth cannot provide us with a defense of our justificatory or verificational strategies: this Truth is the result of their application.
Under the current circumstances, we are looking for the implicit endorsements embedded in methodological commitments. At the most superficial level, it would seem that it is possible, at least logically, to be committed to particular justificatory strategies because they, more often than not, get the story right. That is, it might initially seem possible to make a particular methodological endorsement on the basis of that method’s degree of success in confirming the truth. This analysis of truth, however, suggests that such an endorsement is circular in precisely the wrong sort of way. If truth signifies the successful application of one’s methods, then the above methodological endorsement becomes one based upon the method’s degree of success in being successful. Where such success can make no reference to “truth” (which would only collapse “success” back into some notion of successful justification), we are left without any clear sense of the stake we have in being right, or what it is that we affirm (about ourselves) in satisfying our justificatory obligations.
Something more must be contained in the package of endorsements we undertake by engaging in sophisticated fact language. The task of uncovering this trove is unfortunately complicated by a frequently overlooked feature of implicit inference. Just as an action I thoughtfully perform may have unavoidable (and foreseeable) consequences I nevertheless fail to anticipate, there may be commitments I undertake by pursuing a particular justificatory strategy which I do not appreciate or expect. The fact that “seeking truth” cannot give me a minimal “interest” championed by my justificatory tact suggests that the content of such an underlying endorsement may be fairly complex. The possibility of unnoticed endorsements, lurking in the backwaters of my strategic commitments, further complicates matters.
Recall that we are not concerned with those commitments deeply embedded in my world- view, which are not ingredients of, or relevant to, the strategies I use to defend my fact claims. The only commitments which matter here are those which I must have in order that I believe in the correctness of a given defense. The failure of “seeking truth”, hopelessly underdetermining any tangible commitments with regard to our interests as knowers, and the possibility of ignorance with respect to one’s own epistemological presuppositions, naturally shifts our focus to the justificatory practice in itself. This extends the analysis beyond the conscious interests and beliefs of any particular knower, to capture the systemic endorsements essential to a way of thinking, or a way of arguing. It is, to speak loosely, to treat a justificatory strategy, a scientific method, or even a whole scientific practice as an institution, and to hold the practitioners as partisans to its organizing principle(s).
Where shall we find these organizing principles? As we are seeking the basis of justificatory legitimacy, it would seem natural to investigate the basis upon which a practice or strategy originally gained currency among knowers. That is, if we are interested in the presuppositional grounds for the legitimacy of an appeal to factor analysis, we should look to the way in which factor analysis became an accepted way to argue. Presumably, there would be a great deal of normative residue (including the substantive legitimizing claims) left intact within the practice, even after the practitioners had begun taking the strategy for granted. A style of analysis akin to Michel Foucault’s institutional investigations could then uncover the substance of our unwitting commitments, by prying open the past, and revealing the practice’s ideological (or more generally, political) heritage. The power which legitimated a given justificatory strategy, or the power wielded by the genesis of that strategy, would remain (so the story goes) the substantive basis for its continued vitality and “legitimate” use.
I mention “factor analysis” here because, unlike many justificatory or explanatory strategies — from the complex (eg. use of the calculus and its theory of a limit), to the mundane (eg. Henry’s “show and tell” defense of his bucket’s deformity), it both bears a clear ideological origin, and has since (with increasing frequency) contradicted or abandoned that heritage. The reasons, both for the ideological clarity of its origins, and the subsequent liberation from such conspicuous partisanship, are instructive.
Roughly speaking, factor analysis is a statistical procedure for identifying possible causal links across elements of a multidimensional data set. When collecting biological data, for instance, it is frequently helpful to compare different measures across a common sample rate: so, for instance, you might compare the rate of change in arm length among members of a chimpanzee community (across a four year period), to the rate of change in leg length (across the same period) — particularized to the changes of each individual chimp, of each age group (somehow defined), or of the community as a whole. The actual meaning of each variable in such comparisons is not relevant to the mechanics of the procedure. One measure is mapped onto the other (according to some common reference, like the time of measurement), to generate a two dimensional image indicating the degree of correlation between the two variables. Where the data aggregates toward a line extending from the origin, there is said to be a positive correlation (As the value of one variable increases, the value of the other variable tends to increase). Conversely, where the data aggregates toward a line perpendicular to a line extending from the origin, there is said to be a negative correlation (As the value of one variable increases, the value of the other variable tends to decrease). If there is no tendency one way or the other, there is said to be no correlation. Factor analysis, proper, involves a great deal more than this; but the calculation of “correlation coefficients” lies at the very core of its practice.
What is critical to notice at this stage is that “correlation” has nothing to do with the systemic or causal interaction of real world entities. Certainly, many causal interactions, if measured appropriately, will show up as having high degrees of either positive or negative correlation. But the opposite is not true. In fact, most cases of strong correlation do not indicate any sort of unifying entity or cause. (Unless, of course, the Deist in you wants to maintain that all things are moved by God’s hand. But then I will say that, most cases of strong correlation do not indicate any sort of special entity or cause, which explains this correlation, and the absence of which explains its boundaries.) Nevertheless, factor analysis can be a powerful tool. It reduces the complexity of large data sets, and sometimes suggests places we might look in order to find causal interactions or hidden (second hand) entities.
The fact that the technique does not univocally (or even predominantly) pick out real world properties, and sometimes even misses or conceals them, should raise a troubling question. As much as the method is helpful when you already have some idea about what is indicated by your data, the merit of massaging generic data by factor analysis is not terribly obvious. It would seem strange to devote the energy needed to develop its mechanics, or justify its application, without some fairly clear benefits. There are a myriad of undeveloped methods for massaging data: some of which may be quite useful if we spent the time to work out the intricacies. Why invent this method?
Stephen Jay Gould notes:
I began my career in biology by using factor analysis to study the evolution of a group of fossil reptiles. I was taught the technique as though it had been developed from first principles using pure logic. In fact, virtually all its procedures arose as justifications for particular theories of intelligence. Factor analysis, despite its status as pure deductive mathematics, was invented in a social context, and for definite reasons. And, though its mathematical basis is unassailable, its persistent use as a device for learning about the physical structure of the intellect has been mired in deep conceptual errors from the start.
The “conceptual errors” which “mired” its early application, also brought the method sufficient attention that its mechanics could be thoroughly explored. It was the early application of factor analysis, by Charles Spearman and Cyril Burt, in the realm of intelligence testing, and for the purpose of further demonstrating some innate hierarchy of personal worth on the basis of “intelligence” (with White European males at the top, and Black Africans somewhere near the bottom), that gave the procedure its foothold in general statistical practice. The link with Spearman’s and Burt’s theories of intelligence is not incidental to the method’s development or acceptance. The method was useful, and important, because it gave us one more reason to believe that “we” (the practitioners — generally White, European, and male) possess legitimate power over “them” (the non- practitioners), in the form of being “more intelligent”. The method was legitimate because it could “determine” intelligence in accord with our expectations. The legitimacy of our expectations was not up for grabs; it was this prior legitimacy which lent authority to the statistical analysis which confirmed their message.
But does all this mean that I must endorse the world-view of Cyril Burt or Charles Spearman when I use factory analysis to make rudimentary claims about real world correlations? I hope it is clear that I may not. I may even use factor analysis itself to argue that the link between adult “intelligence” (as measured by any one of your favorite IQ tests) and childhood nutrition is greater than between intelligence and genetic “raw material”. There is, in other words, nothing incoherent in using factor analysis to show that the ideological presuppositions, which buttressed its development and legitimated its entry into statistical analysis at large, are flawed or clearly mistaken. Indeed, we might want to say that, once born, the method is free to confirm or contradict its normative heredity. The institution, for all we may expect, may make an adolescent assault on its founding principles, without fearing that the rejection of those legitimating premises will lead to a collapse of the practice.
Unfortunately, we are not being entirely fair to the motivating concern here. It is not just any aspect of the normative commitments associated with a particular justificatory or explanatory strategy which we are after; but those which are necessary to secure the legitimacy of that strategy. If the legitimacy of factor analysis forever depended upon specific assumptions about intelligence, then it could never have become a tool capable of even addressing non-cognitive matters. The generality of the procedure, as well as the possibility of directing it against these presuppositions, requires that the relevant underpinnings of the strategy either changed after the method gained widespread currency, or were never fully located in a specific account of human intellectual differentiation.
At times an historical analysis will tell us a great deal about the embedded assumptions and primitive commitments which accompany apparently neutral factual analyses. But more often, we must look for the underlying endorsements, which buttress and justify methodological appeals, in the contemporary practice — both of the individual scientist(s), and the modern discipline as a whole. The reason for this is not far to seek. For something to count as true or false [to us], it must make a difference [for us]. And this difference can never be merely methodological. It must challenge or secure the ends we pursue — not only as scientists or knowers, but as persons (i.e., moral actors). Even at the individual level, the outcome must be seen [by me] as relevant to my worth, my status, as a player in the “knowledge” game.
Commitment to a method will rarely reduce to a cluster of non-normative or non-moral commitments. A heuristic device is only as “good” as the virtue of its end; “good at pursuing something”; good at giving us a particular sort of result. I seek an end, and the device facilitates my approach to that end. The merit of such a device can only be as good as the implicit merit of its goal. A method is never merely a method. It is always a method for some purpose or other. And if we are committed to a Tarski-style redundancy theory of truth, the purpose of any method can never be simply “the discovery of truth”.
I can certainly “try one on for size”, without fully endorsing the method’s picture of who we are, or who I am, to be practicing it. But when I apply a method as a justificatory strategy, and expect my fellow practitioners to agree with (or at least consent to) my application, I must be taken to endorse this picture. If I refuse to take myself as being committed to the method’s constitutive premises, one of which articulates our role in the universe, I cannot be taken to be offering an authentic defense of my original assertion. I can only treat a particular justificatory strategy as justificatory if I endorse its picture of who we are to be justified, and what it is like for a creature such as I to be justified, in such a context.
For an analysis which seeks to articulate the particular (normative) content of generic fact claims, an alarming message becomes faintly visible. Particular strategies, like factor analysis or infrared spectroscopy or “inference to the best explanation” (whatever that’s supposed to mean), do not internally contain the material endorsements necessary for their own legitimacy. Indeed, science is so vibrant and diverse a practice (in part) precisely because it does not present a unified core in this regard. Scientific practice does not speak with one voice, either about what we ought to be interested in with respect to our world and each other, nor about what may stand as evidence that we have “gotten the story right” in those same respects.
Building bridges, predicting earthquakes, and successful advertising may depend upon a high degree of coherence in all these regards, so long as we remain within each field or subdiscipline. Movement between engineering, geology, and “market research” will certainly involve a jarring realization that some basic premises about how we ought to think of ourselves may differ substantially, even though we think we are still speaking the “same language”. But even within each discipline or “school”, there will be differences of lighter shades. The danger is that these normative presuppositions will be taken for granted or forgotten. And from there, we might think we have “concluded” something about ourselves which need not be the case. We might imagine that we have no choice but to see ourselves thusly; and bind ourselves unwittingly to a fate of our own making.