A Posteriori Necessity: The Status of Phenomenal Judgments Within Humean Skepticism

Humean skepticism, in a one reasonable analysis, denotes the inability to demonstratively prove propositions which are experientially derived. For Hume, if we cannot demonstratively prove a given set of propositions and the relations they exhibit, then there is no knowable necessity for that set. Knowledge, as it is instantiated in our experiences, is at best probabilistic. Certainly, it would be grossly inaccurate to attribute a kind of favoritism to this doctrine; Hume in no way said that demonstrative (e.g. geometric and arithmetic) knowledge was in any way better than probabilistic knowledge. His skepticism seems to be a result of his realization that metaphysics was grossly overestimating its own power in regard to finding necessary truth in our experiences. As students of Hume realize, it is an extremely lofty pursuit to find apparent fallacies in his principles of human understanding; he seems to be one of the more eloquent and careful writers of the modern era, and it is completely appropriate, I think, to realize a strong resemblance between his brand of cautious and discrete treatment and the thorough and critical manner of modern science. One with most any basic understanding of the modern era and the rise of science should realize that science has, in many ways, adopted Humean skepticism. The unification of math and science, it seems, was at least partly an effect of Hume, rightly so, pointing out that if we are to find a kind of truth in our experiences, it is of a probabilistic sort. However, I believe to have found some fault in Humean skepticism as it applies to a very strict class of experientially-based propositions. And I should like to inquire as to whether or not it is at all possible to have propositions of a posteriori necessity.

Put another way, is it possible to know anything paralleled or accompanied by experience to escape the Humean claim that my knowledge of that proposition is probabilistic, and not necessary? While we would all love to think of our inquiry as truly revolutionary, this seems an impossible reality for philosophy in its current state; others before me have articulated similar notions, yet I purpose a different way of thinking about the possibility of a posteriori necessity, in light of semantic and contextual differences contained within (perhaps) the most basic of words relating existential import, 'is'. In light of this realization, my pursuit is to show that a certain kind of experientially-derived proposition manifests a level of certainty chiefly distinct from that exhibited by the propositions open to the skepticism Hume was treating in his Treatise of Human Nature and Enquiry Concerning Human Understanding. Whether or not we may say that this new class of epistemic claims presents the same level of definiteness Hume attributed to mathematical, a priori necessity remains to be seen, and it remains only speculative insofar as what this inquiry is to acknowledge.

Insofar as what Hume is treating, I believe the skepticism concerning the contingency of experiential and/or empirical knowledge a tenable and probably irrefutable conclusion. While this doctrine holds in most scenarios, it does not hold in others that, nonetheless, are necessitated by experience. We shall find that some types of propositions can be said to be necessary, even if they are experiential and thus a posteriori. If this claim turns out to be correct, as I endeavor to show, it seems Hume's epistemology is, in one limited but important application, fallible. More specifically, Hume's skepticism is appropriate only for experientially dependent propositions in which 'is' denotes a description. However, in analyzing the relatively infrequent cases in which 'is' denotes an identity relation to the judgment of consciousness upon itself, a kind of awareness about experiential-awareness, we see a breakdown of the notion that all experientially derived propositions are merely contingent. However, we cannot articulate further unless we have a grasp of the relevant propositions Hume illuminates in his epistemological doctrine.

Hume begins his doctrine of human understanding by delineating between two kinds of objects of human reason: relations of ideas and matters of fact.[1] Relations of thoughts, we are told, signify those propositions which exhibit conceptual truth[2] and thus are known by the "mere operation of thought, without dependence on what is anywhere existent in the universe". This echoes what students of logic are told in an introductory course: universally quantified propositions are merely relations of ideas, and lack the existential import of particularly quantified statements, which express at the very least that something exists. Professors tell us to be wary of universals, as universals attempt to relate an entire class of entities with another class, and once the hypothetical is projected upon the actual, it seems large classes do not operate in this simplistic way. Thus, it is tremendously difficult to ascertain soundness in this way. And yet, we witness a rather ideal implication out of the simplicity of a priori knowledge: if propositions of this kind are true, they are certainly true. Likewise, a priori propositions that are demonstratively false are certainly false. The binary relation between truth and falsity may surely be the most attractive feature of mathematics and deductive logics, for they are justified in saying 'p is of certain truth in virtue of the concepts involved'. A post-Hume scientist speaking in a similar fashion would indeed be ridiculous. This is the case, for Hume, due to the very different truth conditions we use to analyze the truth or falsity of matters of fact.

Matters of fact are the second objects presentable to the human mind, and the conditions under which we are said to know objects of this type do not employ the binary (and absolute) relationship between truth and falsity that is so inherent to those objects of the relations of ideas. We may briefly summarize that, concerning matters of fact, Hume exposes the contingency of propositions that are derived from experience. He writes:

"The contrary of every fact is still possible; because it can never imply a contradiction, and is conceived by the mind with the same facility and distinctness, as if ever so conformable to reality. That the sun will not rise tomorrow is no less intelligible a proposition, and implies no more contradiction, than the affirmation, that it will rise."[3]

Matters of fact are ultimately rooted in experience, and our experience of something in the past need not represent the future, which may or may not present similar circumstances. If we cannot prove a priori that the past will even resemble the future, we cannot a priori deduce the necessity of the relationships of our perceptions. We may observe, and we may experiment, and we may encompass a very large class of experiences for any one given phenomenon we wish to inquire about, but we have no grounds for absolute certainty, for the fact that in the past I observed all A's to be B's in no way necessarily implies that the conditions under which I posited that relation will be as such in the future. Thus, when we wish to acquire knowledge as to things we experience, our method of evaluating truth is probabilistic, and not demonstrative. We see a change in the conception of truth and falsity from a binary relationship to one that implies degrees as we move from the a priori to the a posteriori, for we are said to be 'more compelled' to believe something we experience in subsequent instances; it is a rather natural inclination of the minds of both animals and humans, and it is one which functions to increase our chances, a coping mechanism of sorts.

In this great instance, we see in Hume the reasons why psychology and evolutionary biology are feasible, for the nature of man is quantitatively, not qualitatively distinct from animals. It is rather an ironic occurrence, for post-Hume philosophy no longer enjoys that inflated conception of itself, and perhaps it realized, so to speak, that wisdom and truth were harder to get at than originally proposed, when the work of metaphysics posited time and time again that it had found the whole truth regarding the nature of all things. Although not directly relevant to the purpose at hand, it is a reality the inquisitive and insightful must accept.

I suppose, in Hume, we have a sort of inverse relationship between absolute certainty and existential import. On the one hand, propositions of the relations of ideas inherently claim an absolute certainty if true, yet they wholly lack any existential quality. On the other hand, matters of fact, while devoid of absolute certainty, predicate existence insofar as we say that our experience (our perceptions) instantiates some kind of existence. This is all present in Hume so he may move to the all-encompassing defeat of metaphysical causality. Well, perhaps defeat is not the appropriate term, but at the very least, we may say that within a very few pages, Hume succinctly illustrates the ridiculousness of attempting to ask questions which endeavor and in some very broad sense, necessitate the appropriation of causality within metaphysics. Unlike the critiques of his predecessors, what was problematic for Hume were not the solutions produced by the metaphysicians (e.g., Descartes, Spinoza), it was the questions they were asking, and the underlying assumption that we could arrive at some analytic, self-evident truth even when we are referring to those things which are completely experiential.

Let us now briefly provide a synopsis of the relevant findings we have developed: 1) geometric and arithmetic propositions yield a kind of binary relation between truth and falsehood which is instantiated through a kind of analytic necessity; that is, their truth is one which is strictly relational through the conceptual framework they evoke; 2) experiential propositions are ultimately dependent on simple impressions and their correspondent simple ideas, which differ only in magnitude upon the mind; that they are contingent and probabilistic affords a kind of gradualism with respect to the notions of truth and falsehood, that is, one is convinced more of the truth of a proposition when, under similar conditions, that proposition affords a experientially supported claim. Thus, in conclusion; Humean skepticism in one understanding requires us to be always, even if slightly, skeptical of propositions derived from experience, as the past may not conform to the future, and that we have not the grounds to found the formation of experiential propositions within the iron-clad certainty of necessity. In effect, a posteriori propositions can not yield necessary relations. It is this last point with which I should like to argue, yet it shall require a more thorough classification of the kinds of propositions our minds are presented with.

A proposition shall be understood, generally, as a judgment upon a certain class of conceptual relations that functions to fix a set of conditions so as to make the relations between concepts comprehensible. By comprehensible I mean that, given a realization of the accessibility relation between concepts of the propositions, a valuation of truth is derivable. If we were to bring this characterization of a proposition to a realistic setting, we make the following example of what Hume characterizes as an a posteriori contingent claim:

P©û: That ball is red.

Let us decode P in light of this characterization of a proposition. If P is a proposition, we may say it is a judgment. What now, indicates the nature of the relation between the relevant concepts here of ball and red? Surely, it is the most basic of existential relations, 'is'. 'Is' here denotes an existence of description, since one way to characterize, describe, and understand a ball is by its color, but of course there are other ways. Thus, it seems that we may arrive at a conception of the descriptive use of the relation 'is'. The 'is' of description yields a relation between two or more concepts that manifests one comprehensive way of understanding the functioning between two or more concepts of a given proposition. There is something else about this proposition though and we should have to discuss this first and foremost before returning to where Humean skepticism is inapplicable. This proposition is telling of a kind of correspondence between idea and impression central to the nature of our minds. Hume of course discussed this: when we have a collection of simple ideas contained within an assertion, there is a corresponding and similar collection of simple impressions. This construal is rather evident in the following: "...all our ideas, or weak perceptions, are derived from our impressions or strong perceptions..." He explicates further, "...that impressions always take the precedency of them, and that every idea with which the imagination is furnished first makes its appearance in a correspondent impression."[4]

What is acknowledgeable in scientific study is the correspondence between impression and idea. Impressions are physically manifested as shown through various tools of neurology, including MRI and PET scans, for instance. Thus, when someone is having an idea, he or she is manifesting the idea via the physical instantiation, which is the corresponding impression. I know this is not the conventional way of understanding this relationship in Hume, but I do think it is at least reasonable, and implies something else which answers the question of how exactly Humean skepticism is to be regarded presently in say, the science of the mind and behavior.

Simply put, Humean skepticism is very much applicable to the practice of the psychological sciences, given that the inquiry is to this relationship between idea and correspondent impression. Impressions, in this context, are discoverable by means of experience: we take measurements with fMRI's, and PET scans, we look at exactly how a person's brain is reacting when the person is confronted with a set of physical stimuli. Thus, it is plainly see that, given any proposition that is experientially based insofar as there is a correspondence between simple idea and simple impression, we may say that Humean skepticism still applies. Theories of psychology are still based on probabilities in the form of, for instance: "in the past we have noticed that when a subject is presented with stimulus or task X, his or her brain has, as indicated by our data, increased in metabolic rates in this area, but not this area...", and so forth. This kind of knowledge is very much probabilistic, so we find no fault or inapplicability for Hume's skepticism.

Returning to how we may characterize P, let us say that it is, as philosopher David Chalmers has recently provided, a first-order phenomenal judgment.[5] Put simply, any proposition is a first-order phenomenal judgment if it parallels awareness, but does not refer to it.[6] In Humean terms, first-order phenomenal judgments at the very least cohere with his notion of simple ideas. Recall that simple ideas may be construed as propositions which are experientially derived in that they correspond to and are formed from simple impressions; it is not the case that simple ideas are concerned with conscious experience, but merely parallel it, or rather are necessitated by it. And so, we may rightly say that first-order phenomenal judgments in some way are instantiated in Hume's doctrine regarding simple ideas. Thus, first-order phenomenal judgments are open to skeptical doubt. These are the propositions which may serve as evidence for or against any scientific theory, I should wager.

Finally, I think a relevant understanding of Humean skepticism has been provided, and thus it is appropriate to probe into the inappropriateness of the skeptical argument. Consider the following proposition, one that in some sense is necessitated by experience:

P©ü: My sensation is red.

The difference between the first proposition and this second one is perhaps not as significant as the initial reading of the thesis may have implied, but the referent of P©ü is concerned with conscious experiences themselves, thus we differentiate between the two by calling this kind of proposition a second-order phenomenal judgment.[7] This second proposition significantly contrasts with the ultimate referent of first-order phenomenal judgments, which refer to objects of sensation (essentially classified in Hume's matters of fact). The semantic nature of this proposition disallows such a proposition to be functionally interpreted; readily, the reader will admit that no such proposition would find its way into the data of a certain experiment that perhaps was probing into one kind of perception. Indeed, the reader may argue that P©ü is physically correlated, and this is because someone uttering this proposition would have physical correlates showing up on a brain scan. This, as such, is the case; however this response fails to deal with the heart of the matter. If we look at the semantics of P©ü we realize that consciousness is reflecting on itself. In effect, the perceiver is reflecting not on what he takes to be existent in sensory datum[8], but upon the nature of that which allows him to be aware of things.

Second-order phenomenal judgments are not reducible to scientific analysis, because even if they are physically correlated in terms of what a brain scan may indicative, their semantic necessity is one which is derived from experience but does not require experience to be known to be true. They need not be of this particularity either; Chalmers mentions a third-order[9] which are essentially characterized as more general forms of second-order judgments, but the distinction seems to be rather weak, and the heart of the matter remains that first-order judgments parallel awareness, while second-order judgments refer to them.

In no way do I claim that propositions of the second-order are true. I am merely claiming that if second-order propositions are true in any given instance, they are necessarily true in virtue of the relations of the concepts they maintain. Consider Descartes' res cogitans (C), which I believe, serves as a rather prime example of the second-order judgment:

C: I am a thinking thing.

The problem Hume's doctrine of human understanding fails to account for is the necessity of the truth conditions. It is not the case that we are to experience a given set of phenomenon which indicates to us that a mind is that which thinks; that would be a ridiculous practice. No, C implies that C is necessary in the moment of the conception; it is a claim about consciousness and its nature, a kind of reflexive 'is' which denotes an identity relation between consciousness, which we may roughly construe as an innate awareness of mental and physical events, and the conscious perception of an irreducible character of consciousness. Many will find fault my claim that our awareness of the fact that we think is not functionally explainable, and I welcome critics, however there is neither space nor time to delve into the matter too much[10].

Propositions like C are valuated true under very different conditions than what Hume accounts for in the contingency and probabilistic character of experiential knowledge, and yet we readily acknowledge propositions like C are experientially-derived. In that I mean, a very young child, devoid of sufficient experience of the world and of her internal thoughts, would not have the capacity to make a claim like C. In light of this, suppose we were to claim that, although second-order phenomenal judgments are derived from experience, they not of the same sort. I know we should like to maintain that all experience is of one kind, and that is in the realization of, in the end, simple impressions and their corresponding simple ideas; however this skeptical position cannot tolerate second-order phenomenal judgments, which if true, are true in that binary sense that necessity infers. I should like to add that I doubt anyone in their sane mind would seriously doubt the truth of C, which perhaps serves some supportive function to my thesis and the need to divide experientially-based propositions into two kinds reflecting a difference of truth conditions.

To sum up then let us provide a revision of Humean skepticism. After the above analysis, and the failure of Hume's doctrine of human understanding to account for the binary nature of the truth or falsity of second-order judgments, we may rewrite Hume's doctrine as the following:

H©ü: All propositions that parallel awareness are a posteriori, contingent claims, and thus probabilistic. The 'is' of these propositions yield a relation that is descriptive. These propositions are derived and proven from experience.

For the purpose of clarity, we may revisit the necessity which Hume defined to be manifest in his doctrine of the objects of the mind that are relations of ideas:

T: All propositions which may be demonstratively true are a priori necessary claims, thus the relation of truth and falsity is completely binary. If a proposition of this kind is false, it is certainly false.

Finally, we may posit the characteristics of what may at least be possible in regard to a rejection of the idea that all experientially derived propositions were probabilistically true.

A: All propositions that essentially are about conscious experience rather than parallel to it may be characterized as a posteriori necessary claims. The 'is' of these propositions denotes an identity relation. These propositions are derived from experience, but are not proven by it.

If we take these claims to be true, it seems that those propositions which in some are derived from experience should be broken up into two kinds, one of which is determined true in the Humean sense (via probabilities), and another which is determined true in the necessary sense (via conceptual relations). I realize that Hume's work did not explicitly concern science however it is clear that the skepticism inherent to its conclusion is indeed very much a part of modern science. Furthermore, I might add that Hume's mistake was in generalizing faculties of reason as either to be based on experience or not based on experience, and it is a problem seen frequently elsewhere, in virtually all works that are said to be supportive of either rationalism or empiricism. That debate is one which Hume could not escape given the circumstances of his life, but it is one which no longer is as pressing a concern for philosophers and scientists nowadays because we realize that the distinction is a silly one: of course all experience accompanies knowledge, however there are some things we learn which are so analytically true, like that two plus two equals four, that subsequent positive instances do not make the conclusion more true in our heads. This is the heart of the matter, and one which is more complex for it implies a difference of truth conditions, which is something I tried to inquire about in my analysis. Thus, my rejection of Humean skepticism as it pertains to one class of propositions in no way means I think he was wrong; to the contrary, I think he was right given what his treatise dealt with, I meant only to analyze it within a current understanding of related phenomena.

Bibliography

Chalmers, David, The Conscious Mind: In Search of a Fundamental Theory (Oxford: Oxford University Press, 1996)

Hume, David, An Inquiry Concerning Human Understanding (Chicago: Open Court Publishing Company, 1996)

[1] David Hume, An Inquiry Concerning Human Understanding (Chicago: Open Court Publishing Company, 1996), 71.

[2] Conceptual truth, in this context, is to be understood as the conventional notion of a priori necessity. Both denote a proposition which is analytically true, or true in virtue of the relations of the concepts.

[3] Hume, 71

[4] Hume, 32

[5] David Chalmers, The Conscious Mind: In Search of a Fundamental Theory (Oxford: Oxford University Press, 1996), 175

[6] I opted to use his same terminology because it is a rather novel concept and to rename it under my own characterization would be disrespectful and unreasonable. Its meaning perhaps can be understood better in Chalmers' own terms, though I think my translation of it is sufficient for our purposes. He writes, "First-order phenomenal judgments are the judgments that go along with conscious experiences, concerning not the experience itself, but the object of experience". Chalmers, 175

[7] Chalmers writes that second-order judgments are "...more straightforwardly judgments about conscious experiences." He arrives at a general rule positing that, "it seems that for any conscious experience, if one possesses the relevant conceptual resources, then one at least has the capacity to judge that one is having that experience." (Chalmers, 176)

[8] See Bertrand Russell's theory of neutral monism.

[9] Chalmers, 176

[10] Critics will site many reasons why to not believe that consciousness is an irreducible phenomenon cognitive science will one day conquer. I should like them to realize though at least where I am coming from in this claim. While the psychological sciences are probing into instantiations of consciousness, that is, properties and realizations of consciousness (especially as they may be functionally represented and neurologically based), I argue that acknowledging properties of things is a different endeavor from realizing the intrinsic nature of the phenomenon itself. Put another way, there is a difference between accounting for the physical correlates of, say first-order phenomenal judgments, which I agree do occur in consciousness, and explaining the explanatory irrelevance of consciousness. This is evident in the fact that we have the ability to construct propositions like C, and yet all our psychological theories regarding the causal relations correlating with C refer not to consciousness itself, but only the accompanying perceptual processes C manifests in a brain scan. I ask, if phenomenal judgments are on some level based on the physical laws of science, given that we can utter such propositions, then science should tell us one day why we are disposed to make judgments about consciousness. However, it seems only to be able to account for the more Humean question of why we are disposed to make judgments that merely parallel experience. Why consciousness is also able to make judgments regarding its own awareness is puzzling to the scientist even today, and remains a largely unresolved issue. There is a critical distinction between those two endeavors, and the former provides insight into the very nature of consciousness itself, irregardless of its experiential basis.