Knowing Way Too Much: A Case Against Semantic Phenomenology
© Springer International Publishing Switzerland 2015Michał Araszkiewicz, Paweł Banaś, Tomasz Gizbert-Studnicki and Krzysztof Płeszka (eds.)Problems of Normativity, Rules and Rule-FollowingLaw and Philosophy Library11110.1007/978-3-319-09375-8_4
4. Knowing Way Too Much: A Case Against Semantic Phenomenology
Institute of Philosophy, Jagiellonian University, Kraków, Poland
Proponents of so called “semantic phenomenology” claim that we are able to hear meanings when we hear meaningful utterances. Recently, Philip Goff has proposed his interpretation of semantic phenomenology that leads to the conclusion that the sceptical problem posed by Saul Kripke in his Wittgenstein on rules and private language must be solvable. My aim in this chapter is to question this view by showing that the way Goff conceives the epistemology of meanings is not compatible with the basic intuition about the possibility of linguistic error. Consequently, we cannot rightly say that our phenomenal experiences represent meanings. The conclusion is that the existence of the conscious phenomena described by semantic phenomenology is irrelevant to the ontological problem of existence of rules and meanings. At the end, I sketch an alternative picture of the role played by conscious experience in our use of language.
Proponents of semantic phenomenology claim that we are able to literally hear meanings while listening to meaningful utterances we understand (Goff 2012; Siegel 2006). The primary reason given for this claim is that there is a certain phenomenal difference between the situation when one hears an utterance and understands it and the situation when one hears words that one doesn’t understand. The very idea of semantic phenomenology is subject to criticism (see e.g. O’Callaghan 2011). For the present purposes, I will however assume that we can speak of such a thing as experiencing a linguistic expression as meaningful. What will be contested, on the other hand, is what consequences can be drawn from that thesis.
Philip Goff has proposed his version of semantic phenomenology (Goff 2012), which leads to the conclusion which is at odds with scepticism about rule-following. My aim in this chapter is to show that even if we agree that semantic phenomenology exists, it doesn’t lead to any definite conclusions about the solvability of Kripkenstein’s sceptical problem.
Let me briefly state the rule-following issue. As it’s well known, Kripke introduces his sceptical problem using the following famous example (1982, p. 9): imagine that 56 is the biggest number you’ve ever used while performing addition (of course this is very improbable, but for any person an “upper limit” of his addition can be found). Now, you are asked to perform the operation of “57 + 68”. Naturally, your answer is “125”. But then you are confronted with the hypothesis that what you really mean by “+” is not a familiar addition function, but rather a quaddition function, which returns the same values as the original addition function for arguments that are both lower than 57 but 5 for all the other arguments. Kripke-Wittgenstein’s sceptic is then to show that confronted with such a hypothesis we are unable to provide a satisfying answer to her challenge. There is nothing we can invoke in order to support our belief that we mean the “right” function.
The story is familiar enough so I won’t elaborate it any further. It’s important to keep two things in mind. First, the argument is meant to be strictly general—the sceptical conclusion applies to any meaningful use of language, not just arithmetics. Secondly, the negative conclusion is understood in ontological, not epistemological terms. It is not that we lack cognitive capacities to recognize some fact that makes it true that by “+” we mean addition, but that there is no such fact (Kripke 1982, p. 39). It is worth noting, however, that the negative epistemological thesis follows from the ontological one—if there is no fact of the matter regarding the meaning, then we are in no position to know it. Even an omniscient God would not be in a better position. Still, one should bear in mind that the ontological claim is more fundamental.
In my view, the way Goff conceives semantic phenomenology leads to the conclusion that the very fact of its existence is enough to show the solvability of the Kripknesteinian problem.
When we consciously perceive someone using an assertoric sentence in a language we understand, our perceptual experience represents the speaker as making a certain claim; to return to the example, if I say to you ‘God is a friend to all’, your perceptual experience represents the utterance of that sentence as an act of literally claiming that God is a friend to all. This perceptual experience involves hearing the words in that sentence as meaningful, and this in turn is a matter of experience representing those words as contributing to the making of the claim in specific ways. Your experience represents the word ‘God’ as determining that God is the subject of the claim. Your experience represents the predicate ‘is a friend to all’ as determining that the utterance claims of the subject of the sentence that it is a friend to all (Goff 2012, p. 225).
The key thing about Goff’s version of semantic phenomenology, as is clearly seen from the above quote, is that we perceive utterances not only as meaningful but also as having specific meanings. This is extremely important in the context of the sceptical debate. If we hear “+”, not just as meaning something (as opposed to meaningless), but as meaning specifically addition, then, supposedly, we have a straightforward answer to the sceptic. We know that by “+” we mean addition just because we perceive it.
It is also instructive to look at the way Goff treats his version of the famous Mary example. His story runs as follows: Mary, a genial scientist is ask to determine what a person (called Cuthbert) means by a given expression. When Mary is given only physical information, regarding her subject dispositions etc., she is unable to determine whether Cuthbert is experiencing “plus” as plus or quus. But the task becomes simple when she is provided (by an evil daemon) with a glimpse of the subject’s phenomenal states:
instantly Mary knew everything there was to know about what it’s like to be Cuthbert at t. She now had access to Cuthbert’s semantic phenomenology at t; to what meaning Cuthbert’s perceptual experience represented the word he was hearing at t as having. Mary had only to reflect for a moment before declaring, ‘Cuthbert experiences ‘plus’ to mean plus’ (Goff 2012, p. 231).
The primary aim of Goff’s argument is to discredit physicalism, but this aspect will be left aside (Alex Miller and Ali Saboohi doubt whether Goff’s paper poses a problem specific to physicalism (see: Miller and Saboohi forthcoming)). What will be the subject of interest is whether the answer to sceptical problem, which is more or less implicit in Goff’s argument, is a successful one.
There might arise a more general worry that Goff’s version of semantic phenomenology is not in the business of providing a direct answer to Kripke’s sceptic. Nowhere does Goff suggest that the phenomenal states in question might serve as facts that make sentences about meaning true.
But let us remember that the negative ontological thesis about meaning-facts implies the negative epistemological thesis: if there are no facts about what we mean then there is no way we can know what we mean. By simple modus tollens we can infer that if there is a way to get to know what we mean (which is precisely what Goff claims) then there are indeed facts about what we mean. So, the very existence of semantic phenomenology disproves the sceptical conclusion, even if we are still in no position to give an account of meaning-facts.
Additionally, as the quotes above show, Goff often speaks of perceptual experiences as representing meanings, and if this talk about representation is to be understood in broadly realistic fashion, then the underlying assumption must be that semantic phenomenology provides us with knowledge that meaning-facts exist. So anyone who is even mildly sympathetic to the view that rule-following paradox shows that the realistic account of meaning and rules is mistaken must deny this interpretation of semantic phenomenology.
While I’m not inclined to claim that it is false that we might have conscious experiences in which we have the impression of perceiving meanings, I doubt whether it’s right to say that such experiences provide us with knowledge of meanings (where knowledge is understood as realistically conceived representation). In order to show that, let us look more closely at the crucial notion of “knowing the meaning of a given statement”. If we say “X knows what P means”, it might be interpreted in at least two ways. First, it might mean that X knows that P means p, for example—Jones might know that “Snow is white” means snow is white. This phenomenon will be labeled, for the purposes of this chapter “disquotational knowledge of meaning”. The basic test of possessing such knowledge is the ability to verbally exclude alternative interpretation of what one says. For example, when confronted with the sceptic, the subject would deny that she means quus by “plus” and insist that she means plus.
Two things might be said about this notion. First, possession of such knowledge is not enough to disprove Kripke’s sceptic negative epistemological thesis. The fact that if someone asks me what I mean by “green” and I reply by saying that I just mean green doesn’t get us any further in the context of rule-following dialectics. To quote John McDowell: