Philosophy 371, Minds & Brains/Cognitive Science Lab
Lab # 2: Mysteries of Meaning
Preamble: The meaning of meaning
Submitted for your consideration: Imagine two sticks, tied together to make an X. Tied with thread to the sticks, a plain jacket button, a sprig of birch, and a sprig of pine.
What is its meaning? "What?!" you say, "this is just a bunch of twigs. It has no meaning, nothing like the meaning of a sentence!" Now read this passage from Robert Louis Stevenson’s Kidnapped:
"So when it comes dark again, I will steal down into that clachan, and set this that I have been miking in the window of a good friend of mine, Joh Breck Maccoll, a bouman of Appin’s."
"With all my heart," says I; "and if he finds it, what is he to think?"
"Well," says Alan, "I wish he were a man of more penetration, for by my troth I am afraid he will make little enough of it! But this is what I have in my mind. This cross is something in the nature of the crosstarrie, or fiery cross, which is the signal of a gathering of our clans; yet he will know well enough the clan is not to rise, for there it is standing in his window, and no word with it. So he will say to himsel’, the clan is not to rise, but there is something. Then he will see my button, and that was Duncan Stewart’s. And then he will say to himsel’, The son of Duncan is in the heather, and has need of me."
"Well," said I, "it may be. But even supposing so, there is a good deal of heather between here and the Forth."
"And that is a very true word," says Alan. "But then John Breck will see the sprig of birch and the sprig of pine; and he will say to himsel’ (if he is a man of any penetration at all, which I misdoubt), Alan will be lying in a wood which is both of pines and birches. Then he will think to himsel’, That is not so very rife hereabout; and then he will come and give us a look up in Corrynakiegh. And if he does not, David, the devil may fly away with him, for what I care; for he will no be worth the salt to his porridge."
"Eh, man," said I, drolling with him a little, "you’re very ingenious! But would it not be simpler for you to write him a few words in black and white?"
"And that is an excellent observe, Mr. Balfour of Shaws," says Alan, drolling with me; "And it would certainly be much simpler for me to write to him, but it would be a sore job for John Breck to read it. He would have to go to school for two-three years; and it’s possible we might be wearied waiting on him."
Meaning, in this case, depends on the interpretation. If a thing (a token) has an interpretation, then it has a meaning, and so meaning can be equated with interpretability. John Haugeland takes this view in Chapter 3, and works very hard to determine what "interpretability" amounts to. His analysis of meaning and interpretability unfolds in several stages:
To interpret is to make sense of. To interpret a system of marks or tokens as symbolic is to make sense of them all by specifying in a regular manner what they mean. The specification typically has two parts: what the simple symbols mean, and how the meanings of complex symbols are determined by their composition (components plus structure). So, loosely speaking, it’s like translating a newly discovered language. You give an interpretation by producing a "translation manual," consisting, roughly, of a dictionary and a grammar. The interpretation that "makes sense" of the symbols -- makes them understandable -- by explaining what they’re equivalent to in some other system (like English) that’s already understood.
The basic principle of all interpretation is coherence: every interpretation must be coherent. There are two sides to the coherence principle.... [Y]ou must start with ordered text, and you must render it so that it makes reasonable sense. (p. 94)
"Ordered text," our author goes on to say, "must have a systematic internal structure that is not accidental.... In addition, an ordered text must have enough internal structure to support interpretation" (pp. 94-95). [That is, there must be enough text given to rule out coincidence and alternative interpretations.]
"Reasonable sense" is Haugeland’s topic for most of the chapter. In brief, he maintains that an interpretation needs to be generally truth-preserving. That means that despite occasional lapses our interpretation scheme turns our untranslated text into an approximately true story or description, with the intimation that if the source of the text produces more text, that the same interpretation scheme will continue to work. One way to guarantee success is to find an interpretation that reads the text as entirely derived from true axioms, using rules of derivation that are truth-preserving. But perhaps this is not the only way.
Tonight’s lab will enable you to explore the mysteries of interpretation for yourself, and thereby develop an independent standpoint from which to evaluate our esteemed author’s view of the meaning of meaning. In addition, the lab will lead you toward the interpretation of the original original meaner, the human mind/brain.
The lab: To fax where no one has faxed before....
Those who assume that there are other intelligent species in the galaxy often imagine that these smart space aliens are eager to talk to us. Perhaps they are, but we should not be too surprised if our first communication from an alien civilization sounds like this:
Hello. You have reached the very advanced civilization on Rigel IV. Your call is important to us. However, no one is available to take your call just now. Please wait for the beep and leave a message stating your name, coordinates, and number of lightyears away. We will return your call as soon as possible. Thank you, and have a nice day! BEEP.
Extraterrestrials, if they really are smart, are probably really busy, and somewhat disorganized as well. They won’t have all day to chat with us, or to answer our space plaques with plaques of their own. Instead, they are sure to communicate with us in the way that all important and intelligent beings communicate, namely, by fax.
The folks at NASA, having reached this conclusion about ETs, have taken their remaining budget for the decade and bought a fax machine dedicated to the search for extraterrestrial intelligence (SETI). The SETI fax is an ordinary machine, but instead of a phone line it’s connected to a radio telescope. The only trouble is that the cosmic background radiation creates a great deal of space noise which results in spurious, meaningless faxes. NASA needs your advice about sorting the real and meaningful faxes from the random ones. Before they fly you to Houston, however, they would like you to practice on a number of terrestrial examples.
With this handout you will receive examples of lines of potentially meaningful text. Some strings are normal human written language. Some are just a collection of signs which may be meaningful individually but do not form sentences or parts of sentences. Some are just meaningless squiggles. With the examples before you, your job is to answer the following questions in writing. Your written answers will be your lab report for this evening.
1. Provide a set of heuristic rules that identify when a string is a candidate for interpretation. Your job here is not to try to decode these strings, but only to say what to look for in the string that suggests to you that it might have meaning (i.e., that the string is in fact part of a sentence in some real human language). The rules you develop will not be air-tight, of course. And the rules may be a loose set, some of them applying to some meaningful strings and some applying to others. Try to be as specific as you can about the "symptoms of meaningfulness." If some rules are more important than others, say so. For each rule, list one or two strings that show where the rule applies, or doesn’t apply.
2. Suggest to NASA a fax that might be sent back to the space aliens to help them interpret English, or to try to provoke them to help us interpret their language. (It may help that you can send both text and images.)
3. Does your experience in this exercise support Haugeland’s views about meaning, challenge them, or suggest additional issues that he does not discuss? Give your reasons for your conclusions.
You should complete your work on this this evening. Discuss your answers to each question, and then divide up the labor of writing it out. You can email it to us, dan.lloyd and jeremy.rosenberg, or deliver hard copy to my office, McCook 325, by tommorow morning.
1. Provide a set of heuristic rules that identify when a string is a candidate for interpretation.
2. Suggest to NASA a fax that might be sent back to the space aliens to help them interpret English, or to try to provoke them to help us interpret their language.
3. Does your experience in this exercise support Haugeland’s views about meaning, challenge them, or suggest additional issues that he does not discuss? Give your reasons for your conclusions.
4. Suppose you had a whole brain scanner, a device that allowed you to "read" the complete state of neural activation for every neuron in your head. Consider a read-out from the scanner, which would be an ongoing record of every neurons’ activity over time. Would the same principles you developed in question 1 apply to this readout. In other words, would you expect the rules you developed in question 1 to indicate that the brain was a candidate for interpretation. If so, why? If not, why not?