Much talk about language has become tangled in a net, one that takes figures of speech to refer to real things.
As a writer, I depend upon figures of speech, but as a thinker I find it is best to know the difference between the figure and the reality. So I try to avoid speech mannerisms that encourage sloppy thinking. Sports broadcasters, for example, will often say something along the lines of he has a lot of quickness instead of he is quick or he moves quickly. This sort of blather helps fill up air time and, after all, it is just a manner of speaking, so I let it pass with only a silent shriek.
Suppose, however, somebody takes the figure of speech literally and decides to look for the quickness in an athlete’s body? In the old days, this effort would be doomed from the start because there is no such physical thing as quickness. An experiment designed to measure the quickness would immediately run into the problem that quickness has no weight and takes up no space. It is just a manner of speaking.
But these days of cyber-reality are different. Now we can say the quickness is encoded in the body and perform experiments designed to tease out that encoding. We might discover that quickness is encoded in the form of muscular elasticity and neural efficiency, so that we can point to a muscle and nerve connection and say, “There, right there is that athlete’s quickness.” And voila, a manner of speaking becomes a thing.
Is that so terrible? Scientists generally try to apply the principle of Occam’s Razor: entities should not be multiplied unnecessarily. In other words, if you can explain an observation by appealing to one thing, then don’t try to use two things. If you can explain quickness by simply saying it is a quality of the athlete’s performance, why make it more complicated by saying the quickness is present in the body?
But wait, says the experimenter, I have found the quickness. It is unscientific to say that something is a mysterious quality instead of a physical fact.
Such experimenters can only escape this linguistic net if they are willing to suppose that all muscles have some variable elasticity and all neurons have some variable efficiency. At a certain point we will observe that a person can perform an action quite quickly. We do not have to add an actual physical quickness into our story. Suppose the experimenters still insist that athleticism depends on quickness and they have found the quickness and denying their discovery uses Occam’s Razor to shave away too much. Then there is not much to be said; the researchers have become entangled in their own ways of speaking.
This “scientific” insistence on the reality of quickness might sound mad, but we see it everyday when we substitute the word “meaning” for quickness. I am reminded of the seeming hopelessness of the tangle by a recent paper in the prestigious journal PNAS, “An architecture for encoding sentence meaning in left mid-superior temporal cortex,” by Steven Frankland and Joshua Greene.
The paper lays bare its entanglement early on, “…theorists have held that the mind encodes sentence-level meaning by…” [p. 11732]. Suppose the paper said, “theorists have held that the athleticism encodes getting-rid-of-the-ball-level quickness by …” We would know the authors were nuts.
Is there such a thing as athleticism? As an abstraction, sure. The word provides a way of speaking about a host of qualities found in certain people. The problem arises when we say “athleticism encodes,” for now we are turning an abstraction into an active agent. Likewise, the mind is an invaluable abstraction that lets us speak about a whole range of experiences, but can we say, as the paper’s authors do, “the mind encodes”? It makes no more sense than saying athleticism encodes. How about the brain? Well, the brain consists of neurons which fire in circuits. The authors probably referred to a mind because, being abstract, it is not so limited.
How about sentence-level meaning? Is that a thing? Philosophers have talked about meaning for so many millennia that it is hard to say no, but the position of this blog is that meaning is a quality that emerges from usage. Like quickness, meaning is real but not reducible to the measurements of space and time. Therefore, it is not to be sought somewhere in the body.
So which of us is right? Me or the authors? Traditionally, disputes are settled by experiment, but experiments are not available in cases where the quarrel is over Occam’s Razor. In these affairs, the test is whether the appeal to more entities is necessary. Can the facts of the author’s case be stated without assuming the existence of a physical meaning?
Lets look at a small sample of the paper’s sentences that take meaning to be a physical thing. I will then suggest in brackets a way of saying the same thing more directly, without all the sports-announcer’s time killing. As I selected my examples a point that struck me was the number of extra abstractions the meaning-is-real argument requires, and how many things a brain must do besides fire neurons along a circuit. I have used red type for the abstractions created by the brain and orange type for brain actions apart from firing neurons in a circuit.
(1) “Human brains flexibly combine the meanings of words to compose structured thoughts.” [Humans combine words to compose sentences.]
(2) “It would be implausible for the brain to allocate individual neurons to represent each possible sentence meaning.” [It would be implausible for the brain to use a different neuron for each possible sentence.]
(3) “instead … the brain employs a system for flexibly combining representations of simpler meanings to compose more complex meanings.” [Instead the brain uses a finite set of neurons in circuits of unbounded variety.]
(4) “…the meaning of a simple sentence is partly represented by filling in these variables with representations of the appropriate semantic components.” [Untranslatable because it is not a statement of fact The authors theorize that the content of “sentence-level meaning” comes from “the values of abstract semantic variables.” Thus, the abstract meaning arises from another abstraction, much as the physicality of quickness might come from an abstract alacrity. Basically, the authors argue that meaning comes from a more general set of meanings, which are also represented in the brain. They are diving deeper down the rabbit hole.]
(5) “…experiments aimed at understanding how the brain… flexibly encodes the meaning of sentences involving an agent… an action… and a patient.” [experiments aimed at mapping the circuit generated when reading sentences involving an agent… an action … and a patient.]
The paper does provide valuable data on changes in brain circuitry when the same words are used in different arrangements: e.g, the dog chased the man vs the man chased the dog. However, the paper’s discussion section is so caught up in the business of encoding meaning that the argument seems essentially worthless. Working with papers like this reminds me of the task of extracting the valuable chemical information from a mass of alchemical mutterings.