[vers. 4] The Solution to the Logocentric Predicament
§1: The logocentric predicament, stated and explained
The logocentric predicament is a predicament not of logic in itself, but of our use of logic. With Henry M. Sheffer, Harvard logician, the logocentric predicament originates and is simply explained,
Just as proof of certain theories in metaphysics is made difficult, if not hopeless, because of the « egocentric » predicament, so the attempt to formulate the foundations of logic is rendered arduous by a corresponding « logocentric » predicament. In order to give an account of logic, we must presuppose and employ logic. (Sheffer, “Principia Mathematica. Whitehead, Alfred North, and Russel, Bertrand,” pp. 227–228)
The issue always arrived at is found in the employment of what one is trying to prove, which, in this case, is the existence of logic and therefore (logically) necessary inferences. Therefore, if there is no use, i.e., employment, there is no predicament. Those means which we use, which we employ, typically, are inferential principles and inferences. Therefore, non-inferential justification gets around the logocentric predicament. Therefore, in order to get past the logocentric predicament, we will have to line out a theory of non-inferential justification that can withstand all skepticism and critique. This is our massive burden to bear, but it is the burden that all philosophers bear, whether they realize it or not.
In light of what we have said thus far, it is clear that non-inferential justification is what we need, and if we can get it, then the logocentric predicament will be no more. Hence, we must understand what justification we are trying for. In respect to the issue at hand, two forms of justification are to be looked at: propositional justification and doxastic justification. On the one hand, if I have propositional justification of logic, then that means I have justification to believe that logic is the case. On the other hand, if I have doxastic justification of logic, then that means I am justified in believing that logic is the case. The difference is that while I could be justified in my belief with propositional justification, it does not mean that I would be justified in my belief, in that, for me to be justified in my belief, I must have a belief. In other words, propositional justification does not require that you believe what you have justified. Where as, doxastic justification requires you to have belief. It could and will be argued that one must have propositional justification before they have doxastic justification. Putting this into clear view, what we have realized is that before one can be justified in their belief, they must have justification. Now that we have the two forms of first-order justification we are to look at, let us look at second-order or meta-justification in general.
In his Defense of Pure Reason, Laurence BonJour says, “The demand for a metajustification is in effect a demand for an overarching premise or principle to the effect that beliefs which are the contents of apparent a priori insights … are likely to be true” (BonJour, In Defense of Pure Reason, p. 143). With BonJour’s words in mind, it can be said that, for him, metajustification is a call for one to justify those things (e.g., in the case of BonJour, rational insights) that one sees as offering first-order justification for their various beliefs. Others have suggested that the demand for metajustification is a demand that one justify their standards for justification. For our purposes, we will follow BonJour’s understanding of metajustifcation simply because the logocentric predicament is itself a call for metajustification. The logocentric predicament is a call for one to justify the principles of logic (e.g., principles of identity and non-contradiction) and inferential principes (e.g., modus ponens). The issue that everyone faces with the call for metajustification is the threat of using those things that one believes confer first-order justification to try and justify a thing that confers second-order justification. Not only is there vicious circularity in this move, as one is trying to justify those things that they believe to of confer first-order justification in the first place, as well as infinite regression, but there is a mistake of trying to use first-order things to justify second-order things, this mistake being an obvious category error. Therefore, in reaction to all that we have put forward, what we desire is non-inferential metajustification.
§2: First part of my response to Fales’ attempt
Many have failed to avoid having to answer the demand of metajustification. For example, Evan Fales, in his Defense of the Given, tries to get around it. His attempt is, as implicated by the first sentence in this paragraph, a failure. First, Fales heads toward non-inferential justification of inferential principles by way of rational insight. I will not comment on the success of rational insight yet, as I will address BonJour who elaborates on the notion more than anyone else. Instead, I’ll just address his specific use of rational insight. So, in regards to his attempt to get out of the risk of infinite regression that accompanies rational insight, he holds that there is no issue with having “an infinite hierarchy of beliefs” as long as it is “provided that I do know that P” (Fales, A Defense of the Given, p. 165). Given that knowledge is justified true belief, or it is for us and our purposes at least, that Fales knows that P by way of rational insight already supposes that he has first-order justification and therefore second-order justification.
The only way he could possibly get out of this issue of supposing both first-order justification (which could possibly be conferred by way of rational insight) and second-order justification (as if he doesn’t have justification for rational insight, then his first-order justification is groundless, therefore meaning he doesn’t know that P because he is ultimately without justification and foundation) is by trying to adopt the moderate rationalist position BonJour describes in his Defense of Pure Reason, and he does no such thing. Now, Fales’ higher failure is not his suppositions pertaining to a priori justification but rather his suppositions pertaining to justification by way of sense experience. Fales takes this general schema of an inference, “Schema (A)” (Fales, A Defense of the Given, p. 151):
(1) B has feature 𝛟.
(2) Beliefs that have feature 𝛟 are highly likely to be true.
(3) Therefore, B is highly likely to be true.
So, we have Schema (A), but before we elaborate on it, let us take note that Fales holds that position that “[s]ince rules of inference can only be demonstrated by being built out of elementary rules of inference, the legitimacy of some rules of inference must be given” (Fales, A Defense of the Given, p. 14). Yet, this obviously goes without justification, and, in fact, side steps the entire logocentric predicament without justification. Onto Schema (A)!
First, Fales recognizes, “an experience, if neither true nor false, has no propositional content,” (Fales, A Defense of the Given, p. 153), “[o]ur model of what it is for one thing to justify another is that of inference (whether deductive or inductive); and surely one thing can be inferred from another only when both things are bearers of truth value — that is, are propositions or closely allied to propositions” (Fales, A Defense of the Given, pp. 152–153), “[e]ither what is foundational, or given, has cognitive (or propositional) content, or it does not” (Fales, A Defense of the Given, p. 153), “[i]f the alleged foundation does have propositional content, then it is possible to see how it might serve as a warrant for nonbasic beliefs; but in that case its claim to be foundational is bogus, since the demand for a metajustification is legitimate” (Fales, A Defense of the GIven, p. 153), “[i]f, conversely, the alleged foundation does not have propositional content, then the demand for a metajustification is illegitimate — but als then this alleged foundation, because non-propositional, cannot serve to warrant beliefs, and so, once again, provides no foundation for empirical knowledge” (Fales, A Defense of the Given, p. 153). Let us take in all these positions Fales is speaking of. Now, he does not take all of these to be the case, but he understands that it is all of these positions in tandem with one another that sets up the dilemma he is to solve. If it is not clear, Fales has laid out the dilemma that when the demand for metajustification arises, and we try to justify it, we can do this two different ways, by proposing a non-propositional foundation or a propositional foundation, and because inference is the only way we justify things, without propositional content, we, therefore, cannot have justification, but if we propose a propositional foundation then we fail to answer the demand of metajustification and enter into infinite regression once more.
Fales identifies the major presuppositions of this objection: 1. “only a proposition — or, at any rate, a bearer of truth-value — can serve to warrant a belief” 2. presupposition 1 “is plausible on the view that a warrant must involve some sort of inference, and that inference can only move from one truth-value bearer to another” (Fales, A Defense of the Given, p. 153). In responding to the first proposition, Fales argues that “it is not true that only proposition-like items can be justificatorily related. As we ordinarily speak, not only beliefs but also actions can be justified” (Fales, A Defense of the Given, p. 154). He furthers, “Not merely the belief that an action is justified, but the action itself, is subject to justificatory considerations … Thus, the notion that justification is a relation that can obtain only between truthbearers is evidently false, since actions are not truth bearers” (Fales, A Defense of the Given, p. 153). Here, Fales is quite obscure with what he means by action itself “is subject to justificatory considerations” (Fales, A Defense of the Given, p. 153). Actions certainly can be justified in the sense of normative justification, i.e., you ought to do action x. But, that actions can be epistemically justified, is not at all, contrary to Fales, prima facie the case. What would it even mean to epistemically justify an action itself? Fales gives the example of sense experience, but obviously, whether or not sense experience can confer justification supposes that action can confer justification (and by action, I believe Fales is not using the conventional praxeological definition of action as intention or purposeful behavior but rather is seeing action as any movement of the subject). Epistemic justification specifically has to do with the justification of propositions, beliefs, etc., in no way can I prima facie recognize how actions such as sense experience fit in there. But, for the sake of argument, let’s grant Fales this premise, even though it goes both unexplained and unwarranted. He then says, “There remains the question whether the justificatory relation between an action and the reasons for it is an inferential one or not. … But nothing really hangs on whether inference is construed broadly enough to embrace those justificatory relations that permit non-truth-value bearing relata, or whether we restrict the usage of ‘infer’ to truth-value-bearing-relata. In the latter case, we must admit the existence of other justificatory relations, ones that do not require truth-value bearers as relata and that, at the same time, may be close analogues of inference” (Fales, A Defense of the Given, p. 154). So, following Fales in his argument, we realize that by, and I’m paraphrasing here, “actions are subject to justificatory conditions in terms of epistemic justification,” means that actions have reasons behind them. But, again, this is obscure. I can have a belief, desire, etc. and pursue its achievement, but I still fail to see how this would not have to do with normative justification rather than epistemic justification. But, in regards to everything else, it still seems obscure, and operating off both “an expansive conception of inference” and the presupposition that action relates somehow to epistemic justification (Fales, A Defense of the Given, p. 154). Now, if we treat this latter presupposition as granted, we clearly have to see the case in which he is trying to recognize it in. This case is sense experience leading to basic beliefs. I want to say, that after he expands on his conception of inference, it is more clear how actions relate to epistemic justification. So, for example, sense experience being epistemically justified in justifying basic beliefs. Nevertheless, the proposition “sense experience justifies basic beliefs” is much different from the action of sense experience itself, and, in this way, it is still yet to be seen how action on its own can be subject to justificatory conditions. Now, Fales holds that “when a belief is justified by sense experience, it has been inferred by those experiences” (Fales, A Defense of the Given, p. 154). This is absolutely problematic for Fales’ position. He says that “justification by sense experience is in fact deeply analogous to that of justification by deductive inference from some other (known) proposition” (Fales, A Defense of the Given, p. 154). He even admits “given the expansive conception of inference that I am adopting, it would no longer be accurate to describe those beliefs that derive immediately from sense experience as noninferential” (Fales, A Defense of the Given, p. 154). This immediately leads to the reentrance of the question of how basic beliefs are justified, i.e., the demand for metajustification. Hence, Fales must do something to get around this, and quick! And, while he certainly tries, I do not believe he succedes at getting around the overall issue, and why I believe this will be explained shortly.
So, in the reentrance of the demand for metajustification, Fales argues that 1. “all empirical knowledge derives ultimately from sense experience” 2. “[s]ense experiences are neither true nor false, and the demand for justification apply to them” (Fales, A Defense of the Given, p. 154). He furthers, “we have a set of beliefs that are derived immediately from sense experience [and t]hese beliefs do require justification must consist in the existence of suitable sense experiences, and a showing that the beliefs in question can be correctly inferred from these” (Fales, A Defense of the Given, pp. 154–155). Now, it is clear that the issue of justifying supposed inferential principles comes up again, I think Fales probably saw this which is why he says, “All derivative empirical beliefs must ultimately be justified by appeal to beliefs of two sorts: beliefs immediately justified by sense experience, and self-evident a priori truths (Fales, A Defense of the Given, p. 155). Here, it is clear that he believes inferential principles such as the one used in Schema (A) (which I have not forgotten about), are some of the aforementioned self-evident a priori truths. But, as said earlier, in order to escape the issue of the demand for metajustification that arises with rational insight, he must adopt the moderate rationalist position, but it is clear that he adopts neither an extreme or moderate rationalist or empiricist perspective. Rather, I believe that what Fales has here is weak empiricism. It is a type of empiricism because he still gives primacy to sense experience.
So, once again, the issues we see for Fales are 1. the issue of metajustification in relation to those inferences from sense experience to basic beliefs 2. how sensory experiences can be justified without a proposition to represent them 3. why sense experiences are neither true or false, which is also the question of why are only propositions capable of being true or false 4. why something neither being true or false means that it no longer needs justification.
So, with this in mind, let us leave these questions to be answered in the part of this section (section two) after this next section where we will go over his take on a priori justification.
§2.25: My response to Fales on the a priori
Fales correctly says that “when the very simplest patterns of inference are being employed, the demand for justification can no longer be satisfied in [the normal] way” (Fales, A Defense of the Given, p. 155). After concluding that “it is easy to see that the analogue of schema (A) [Schema (I), see below] leads to the conclusion that the allegedly basic a priori beliefs (and inference patterns) are not basic after all”, keep in mind though that with Schema (I), “‘I’ names the inference pattern being justified” (Fales, A Defense of the Given, p. 156):
(1’) Inference pattern I has feature 𝛟.
(2’) Inference patterns having feature 𝛟 are (always or usually) valid.
(3’) Therefore, I is (or probably) valid.
After converting Schema (A) to Schema (I), Fales concludes, “[t]he trouble with this metajustification is that its validity depends upon the validity of a (fairly elementary) inference pattern; and if metajustification is demanded for inference patterns generally, it will be demanded of this one (or of simpler ones from which it can be constructed) as well. But any attempt to justify this pattern by means of an argument that conform to Schema (I) would be clearly and viciously circular” (Fales, A Defense of the Given, p. 156). And there we have it, Fales has come upon the logocentric predicament. The logocentric predicament moves Fales to ask the questions (and they are good ones), “Can a priori knowledge have genuine foundations? And, if recognition that some inference is valid can count as foundational, in what does this knowledge of validity consist?” (Fales, A Defense of the Given, p. 156).
So, the issue with Fales’ attempt to answer this is that he turns to Lewis Carroll’s formulation of a proto- version of logocentric predicament found in his “What the Tortoise Said to Achilles.” The reason this is an issue is precisely because Carroll’s understanding, or rather Achilles’ understanding of inference is mistaken (Carroll is not at fault, for Carroll’s tortoise is very smart compared to Carroll’s Achilles) and it is being sourced by Fales. Fales says, “the very act of inferring itself must (at least in the case of elementary inferences) involve a skill or capacity the cognitive value of which is (metaphorically speaking) transparent, so that knowledge of correct performance and of validity does not presuppose some additional and distinct judgment of the truth of a corresponding conditional
§2.5: Second part of my response to Fales’ attempt
§2.75: BonJour’s response to Fales’ attempt
§X: Analyticity, its importance to the solution to the logocentric predicament
Let us take an analytic proposition to be “a proposition that is such that understanding it warrants the belief that it is true” (Flockermann, Epistemic Norms, A Priority, and Self-Knowledge, p. 144). In other words, an analytic proposition is a true proposition in virtue of understanding what it means. But, how can understanding have justificatory force? This question is analogous to asking how analytic a priori intuitions, as the McGrews call them, or rational insights, as BonJour calls them, have justificatory force.
In regards to analytic a priori intuitions, we are to turn to the McGrew’s Internalism and Epistemology. I feel many have and will blatantly misunderstand this text. For example, the McGrews, in an easily misinterpretable passage, say, “But unless and until [people so inept logically that they cannot inuit the validity of modus ponens] develop the ability to recognize the validity of basic rules of inference, logical metatheory must remain for them a closed subject” (McGrew, Internalism and Epistemology, p. 132). Some may charge the McGrews with a dogmatism of a very high degree for this statement. Such a charge, however, radically misunderstands what is being put forward. The McGrews are not saying this as a proof of modus ponens or a priori intuitions. Rather, they are saying this to all those contrarians who would claim they can intuit no such thing. Thankfully, those who hold they cannot intuit the validity of logic and other inferential principles can offer us no objections. And their failure to intuit such things in no way offers a refutation to what we have put forward. Now, if they were to say they intuited that these inferential principles where actually not the case, then we would be in quite the pickle, no? On the contrary, no issue will have risen with such objections for they can in no way extrapolate off such an intuition an objection against us. Let me explain. If I intuit that an inferential principle is justified and true, and then use that inferential principle to infer that I therefore know that inferential principle in that knowledge is true justified belief, then I can say that I have both propositional and doxastic justification in my belief in said inferential principle. The person who has no such intuition cannot say anything other than their intuition. They cannot justifiably infer an objection without an inferential principle. However, could one not argue that once they say they cannot intuit, for example, modus ponens, that we can make the inference for them? All that they are threatening here is our potential loss to the infallibility of our analytic a priori intuitions.
§X: The McGrew’s solution to the logocentric predicament
The McGrews argue that the principles of logic and other inferential principles such as modus ponens “are analytic truths knowable on the basis of an intuitive grasp of the relations among their constitutent concepts” (McGrew, Internalism and Epistemology, p. 105). I prefer to use BonJour’s term of rational insight rather than analytic a priori intuitions, which BonJour, in some of his commentary on C. I. Lewis, inadvertently admits there is no difference between the two notions.
: Epistemologists such as BonJour reject the demand for metajustification as one must use, in the case of BonJour, rational insight in formulating it, therefore meaning the empiricist, in the demand for metajustification, begs the question against the rationalist. This, however, does not solve the issue of metajustification because if the skeptic takes the rationalist position up to then use the rationalist’s tools against him, then we have a predicament. Furthermore, it is a common error in the thinking of epistemologists that the demand for metajustification is an objection, but it can certainly can be a question. In other words, the skeptic can ask, “Why do you need/not need metajustification?” In this case, the rationalist could say that to even put the question forward they would need to use rational insight, but while this may be the case, the rationalist is still avoiding the consequence of the skeptic’s question. Once the skeptic puts forward the question, it becomes autonomous in that one has now thought it. A defeat of the skeptic in dialectical argumentation is not an answer to the question put forward by the skeptic. Now, I understand that BonJour is saying he doesn’t need to answer such a question, but it is clear that in the face of the autonomy of the question of the skeptic, it is clear that he does, or, at least, he does if he wants to be actually justified on both levels of first-order and second-order justification. To expand this further, let us realize that the empiricist’s question begging and the possible question begging of the skpetic in no way guarantees the justification nor the truth of rational intuition. This would be the fallacy of the particularist, which says that because a particular set of behaviors p, in this case rational insight, are necessary to do x, in this case demand metajustification, they are therefore true. We can see that Aristolte, for example, commits this fallacy when he argues that the principle of non-contradiction is true and justified because no one can refute it without using it. Again, just because something cannot be false in no way proves its truth.
: One may then ask, “What is the justification for the metajustification?” This is an important question which this note is designated to address. Firstly, to understand how we answer such a question we must ask another, “Why do we need metajustification, if that thing that confers first-order justification is non-inferential?” The answer is clear. Having non-inferential first-order justification in no way justifies that thing that confers this justification, rather is justifies that which the justification is being non-inferentially conferred to. That this thing itself is justified cannot, without vicious circularity, be justified by way of first-order justification. Hence, the justification of that thing would be second-order or meta-justification. Clearly, this justification has to be non-inferential, or we lapse back into circularity. Thus, when one asks how our attempt at metajustification is justified, we realize that unless it inherently has a justificatory character what we really have is not a metajustification, but rather just a first-order justification that demands a metajustification. In other words, if one has a non-inferentially established first order justification of “2 + 2 = 4” by way of rational insight, for example, what has been established non-inferentially is “2 + 2 = 4.” But, “2 + 2 = 4” only retains its recently established status as justified if rational insight is justified. Therefore, if I gave metajustifcation to rational insight and then someone asked me for a patajustification (here I am using pata- in the sense Alfred Jerry used it as “that which is beyond the beyond” or “meta-meta-”) of this metajustification of rational insight, then we realize that, unless the metajustification justifies itself non-inferentially, non-regressively, and non-circularly, then we don’t actually have metajustification at all. Instead, we just have another first-order justification demanding for an attempt at its own metajustification. I believe Fales puts it better than I can: “First-order justification of nonbasic propositions consists in showing how these are supported by basic ones. Metajustification consists in establishing that we know which propositions can properly be regarded as basic, by showing that the criteria by which these beliefs are selected are indeed truth conducive” (Fales, A Defense of the Given, p. 151).