| One reason for not trying to discuss the current problem in terms of
imaginary worlds or possible worlds and the like is that we often talk about such worlds in a completely assertive, non-hypothetical way: "Sherlock Holmes lived at 221B Baker Street." This is a true statement, asserted as such, without any subjunctive or contrary-to-fact fididddling, yet there never was such a person or such an address in the real world; it is all literally contrary to fact. Worse, within such discussion we can also deal with the real world in a "contrary-to-fact, subjunctive," way: "If Holmes had pursued Jack the Ripper, the gracious lady would have taken back the emerald." What is crucial, then, is not the nature of the world involved, but of what we are doing with it. In one case we are attempting to describe it correctly and we simply assert that it is thus-and-so and face the consequences of being right or wrong. And in the other case, the "subjunctive, contrary-to-fact"? At least some of the time, the act is speculation -- abstraction, extra-or-interpolation. I am not sure that this covers all the cases of sentences that are not now true but are not recommended nor preferred, but it covers many of them. I am also unsure just what "speculation" means in practical terms. Arguments suggest that there are specultive truths and falsehoods, but history suggests that, outside of the hard sciences, where the speculation can be realized, determining which is which is not easy. In general, the rules are rather unclear, moore or less like the rules of interpretation rather than experiment. Indeed one common use for such speculation is to build on an interpretation toward possible experiment. She gives me a look, which I interpret in a certain way. Assuming that that interpretation is correct, then, if I were to do so-and-so, she would do such-and-such. Now, so-and-so is something I can do without too much commitment, so I can now test my interpretation (assuming my speculation is reasonably correct). What is missing -- and always has been except in special cases -- is good rules for when speculation is correct. Outside of the hard sciences, we have to rely on imprecise terms (about a person's character, say), vague and uncertain generalities (about what people of that sort do in certain types of situations -- and [rarely explicit] what situations they will see as of that type) and the like. And these generalizations -- an even the personality types (to stick with that kind of case) -- are themselves often presented as speculative: "If type x is in situation y, he would... ." Some of this may be inherently a closed language game, without a necessary return to the real world, of the sort that logic gives with closing the indirect proof or science gives with an actual experiment. So how do we mark speculation in Lojban? Are there other non-assertive uses not dealt with? |