Sunday, July 09, 2006

The Difference Between Intelligibility and Meaning

I'm going to expand on my last post about the connection between recognition and meaning. In that last post, I wrote that we only name things that are recognizable (if not recognized). There's little point in naming something we would never recognize again.

In general, meaningful propositions look like:
There are some recognizable x1, x2,...xn such that x1~x2~...~xn
where "~" means correlation. For example, "Tim is tall" means that I will correlate the experiences I would recognize as Tim with experiences I would recognize as tallness. Indeed, the correlation in the proposition is itself recognizable because correlations are inherently recognizable. We recognize both the subjects (xi) and the correlation specified by the claim.

I noted in my last post that we've made a mistake whenever we claim that one of the subjects is unrecognizable, since that contradicts our claim that we started by talking about meaningful things.

An unintelligible proposition is one in which we cannot recognize the subjects. For example, we may be told that "Kes qera tovu axia" is a proposition, but not know what it means because we fail to recognize the words as symbols for recognizable things or relationships.

However, a meaningless or nonsensical proposition is one that contradicts itself by claiming that it represents something recognizable and simultaneously claiming that its correlations are unrecognizable. For example, "all apparently uncaused events have undetectable causes," uses the dangerous word "undetectable." Undetectable is safe when used in a context, as in "undetectable with a Geiger counter," or "undetectable with the unaided eye." However, in-principle undetectability is a meaning-killer. We recognize the individual words in their respective contexts. We can recognize an event as a point in time and space where we see a transition of some kind, e.g., a transition in the trajectories of interacting particles. We recognize a cause as a correlation between the initial and final states of an event. We recognize detectability as a correlation of some kind. Yet, we would never recognize apparently uncaused events having inherently undetectable causes. Indeed, this is a prototypical nonsensical proposition - all of its subjects are recognizable, but its meaning is unrecognizable. Recognition and prediction go hand in hand, and that is why non-predictiveness is sign of meaninglessness.

I'll present one more test proposition: "pixies are undetectable with the naked eye." This proposition assumes that the pixies can be seen or recognized using techniques other than the naked eye. Otherwise, we would not know what it is we were talking about. For the proposition to make sense, we must know how to recognize a pixie in some fashion and with some degree of precision. For example, it is enough to say that a pixie is a tiny humanoid with wings. If we saw such a thing, say, through an ultraviolet camera, we would say, "Aha! A pixie!"

2 comments:

Franklin Mason said...

Three short comments:

I understand well enough what you mean when you say that you recongnize Tim and his height. (It seems to me that we recognize them in quite different ways, but that's beside the point here.) But I'm uncertain what you mean when you say that we recongnize correlations. A correlation seems to be an abstract entity and as such (as would have been said by such a one as Descartes) can be recongnized by the intellect alone. Do you admit abstrations that cannot be recognized by the senses? I would have guessed before I read this post that the answer is no.

Second, consider the proposition:

In-principle undetectability is a meaning-killer

I most certainly understand it, but I have no idea what sense a positivist such as yourself could make of talk of the recognition of the objects referenced within it. What is that experience or set of experiences that I recognize as experiences of in-principle undectability. I'm at a loss.

You say: "In general, meaningful propositions look like:There are some recognizable x1, x2,...xn such that x1~x2~...~xnwhere "~" means correlation." How can this account handle a negative proposition such as "There are no four-headed dogs". What gets correlated with what here? Again I'm at a loss.

Doctor Logic said...

Franklin,

A correlation seems to be an abstract entity and as such (as would have been said by such a one as Descartes) can be recongnized by the intellect alone. Do you admit abstractions that cannot be recognized by the senses?

I think the boundary between sense and intellect is too fuzzy to be reliable. I think it's much better to consider experience in general. Hume talks about ideas and impressions and how we might tell them apart. I think this is sort of the approach I take. I experience mental phenomena and physical/sense phenomena. There are differences in that they are generally distinguishable, but distinguishing between mental and physical isn't significantly different from distinguishing between different physical senses, e.g., between sight from sound.

So, yes, I admit mental phenomena in the same way I admit physical phenomena.

What is that experience or set of experiences that I recognize as experiences of in-principle undectability?

I think it is the experience of seeing ordinary language terms put together in a manner that is non-sensical.

How can this account handle a negative proposition such as "There are no four-headed dogs". What gets correlated with what here?

Ooh, good question. I regret using a tilde for correlation when I might now need it for a logical not!

I guess your question is whether "not correlating" is itself a form of correlation. I think that it is. Not recognizing seems as intrinsic a function as recognizing. I know what it means to not correlate a mental image of a duck with the paperclip I see in front of me. I also know what it means to not correlate an image of a four-headed dog with the physical sight of an actual dog that has only one head.