Nearly a decade and a half ago, before logic bewitched me and I fell under her spell, I started off graduate school intending to write a dissertation on something related to philosophy of fiction or fictional discourse (given that that's how specific my dissertation plans were for my first 1-2 years of grad school, I probably should've realized sooner that this was not the topic for me). This year I'm lucky enough to be teaching a 3rd-year undergrad course "Language & Mind" which has reminded me why I was interested in what philosophy can say about fiction, and vice versa, in the first place. For my inaugural contribution to M-Phi, this post will be the first in an (unbounded) series of reflections on what meaning can and cannot be, given the constraints of how language is actually used, both in real and fictional discourse.
It is one thing for a theory of meaning to give an account of simple declarative sentences which are grammatically correct and whose terms refer to existing, uncontroversial objects: "Snow is white" should not be a difficult sentence to analyse if one is to give a theory of meaning of English. It is yet another thing altogether to be able to handle the edge cases, the non-simple, the non-declarative, the non-grammatically correct sentences, the sentences which have non-referring terms, and unfortunately many theories of meaning stumble at these hurdles, providing answers that are hard to swallow. (Note: I am not one who generally thinks that when there is a clash between what a philosophical theory says and what my intuitions say, it is the intuitions that should win. I've been a philosopher long enough to know that my intuitions in some respect are utterly ruined. However, if my philosophical theory entails a conclusion that is at odds with how people think and act about the relevant subject matter, then I do feel entitled to ask my theory to explain why it is there is this discrepancy. In this, I think St. Anselm of Canterbury's approach to the division between logic and grammar was precisely right. To oversimplify it significantly: Grammar is about how people use language, logic is about how people should use language, but more than that, logic should also be able to explain why it is that grammar and logic diverge: Logic should be able to explain why the usus loquendi does not always match the usus proprie.) It is the edge cases that provide the true test for any philosophical theory, and thus it is edge cases that I'll be discussing in this series.
I started off the "Language & Mind" class with two questions, which have become the guiding questions of the class, and three quotes. The questions are:
- What is meaning?
- What are the preconditions for language to have meaning?
And the quotes:
Anadûnê zîrâ hikallaba //
Êphal ê phalak îdôn hi-Akallabêth.
A threigylgweith yd oed yn Arberth, prif lys idaw adyuot yn y uryt ac yn y uedwl uynet y hela.
I expect the average reader of this blog to recognize the language of at least one of these three, but I would be very surprised if anyone knew all three.
Quote 1 is Adûnaic, and translated into English reads:
Numenor the beloved, she fell down //
Far, far away now is She-that-hath-fallen.
Adûnaic is one of the languages created by J.R.R. Tolkien, the most full account of it appearing in "Lowdham's Report on the Adunaic Language", in Sauron Defeated, ed. Christopher Tolkien, p. 413-440. (An overview of the language can be found here). Tolkien's invented languages are well-known for the attention to detail and realistic grammatical and phonological structures that they have, unlike many other fictional languages which are made up in a piecemeal fashion without any attempt to make them mirror non-fictional languages in structure or complexity.
Quote 2 is Welsh, and translated into English reads:
Once upon a time he [Pwyll] was at Arberth, a chief court of his, and he was seized by the thought and the desire to go hunting.
This is the opening line of the story of Pwyll, Prince of Dyfedd, in the Mabinogion, a cycle of prose literature compiled in the 12th-13th C from oral tales.
I chose these three quotes because each of them places different constraints on what meaningfulness can be, where it can come from, and how we must account for it.
The Adûnaic and Welsh quotes are clearly meaningful, as it is possible to translate them into meaningful sentences in English which can be understood even if the original quotes could not be. The status of the quote in Linear A is less clear: It could be argued (and indeed, students in my class did so!) both that Linear A, given that there is no one alive who can understand or decipher it, is therefore meaningless, and that if it were to be deciphered, then it would regain its previous meaningfulness; or it could be argued (and again, I had students willing to take up this side) that it is meaningful, even in the absence of anyone who can understand that meaning, and thus meaning is something which is intrinsic to a language itself, and not dependent on the people who use the language.
However, while Adûnaic and Welsh are certainly on the one hand opposed from Linear A, on other hand they are opposed from each other. Adûnaic, as a constructed rather than natural language, has a definitive moment of creation or inception, and even if it evolved as it was developed, its development is still governed by the arbitration of a single person. Now that that person is dead, that standard of arbitration is gone: There are questions about Adûnaic that are left essentially unanswerable, questions of both vocabulary, grammar, and pronunciation. Welsh, on the other hand, did not have its birth at the hands of a single person, and as a result, there is a standard that can be appealed to for arbitration, whether this be the sum of its uses in the medieval period (if it is Old or Middle Welsh that is of interest), its use amongst Welsh speakers today, the proclamations of some canonical language academy (Welsh doesn't have one; but French and other languages do). No one single person has the authority to say what is meaningful and correct and what is not, and yet these questions can still be answered, unlike the case of Adûnaic.
The lesson in this post will be short and simple, since the post itself has gotten rather long, and it is this: The varieties of language which a theory of meaning must account for is perhaps broader and more diverse than people who are used to thinking of what it means for an English sentence to be meaningful are aware. In a future post (perhaps the next one), we'll look closer at the case of Adûnaic, and the problems that a truth-conditional theory of semantics would face in accounting for the (apparent) meaningfulness of that language.
© 2015, Sara L. Uckelman.