I've been struggling with the terms entail, implicate and infer lately. There's good empirical reason for this, and I hope to post on this at some stage. But for now, I'm going to go back to undergraduate semantics briefly, and try to re-learn a few basics. Please feel free to participate in my on-going education!
Despite remembering these terms from earlier study, I realised that my understanding of them was really pretty shaky when it came to actually needing to apply them, and distinguish between them.
With the trusty help of philosopher extraordinaire Bulanjdjan Maïa, I got to the point of:
An implicature is stronger than an inference. An implicature necessarily follows on logical grounds, while an inference can be derived by reasoning.
Which left me wondering, 'well, what's the difference between implicature and entailment?'
And in looking around for some clarity on the matter (= going back to old textbooks), I didn't find much illuminating material. For example, one textbook says:
Sometimes knowing the truth of one sentence entails or necessarily implies the truth of another sentence.
So, entailment = implicature?? Some further thinking on this got me to:
Perhaps implicature operates at the level of pragmatics, while entailment operates at the level of semantics?
Fortunately some good old Gricean know-how kicked in, courtesy of Michael Haugh's paper on The Intuitive Basis of Implicature. According to Grice (1967, 1989), there is conventional implicature, and conversational implicature. Conventional implicature operates at the semantic level, and entailment is a kind of conventional implicature. Conversational implicature operates at the pragmatic level, and can be defeased.
So that is where I rest my investigation of these terms (for my purposes) for now. However, I was struck by the use if the prefix neo- in Haugh's paper:
Most neo-Griceans have essentially retained this definition of implicature in subsequent developments of Gricean theory, so implicature has continued to be defined as what is communicated less what is said. The problem for the Gricean (and neo-Gricean) definition of implicature is that it encompasses far too large and diverse a range of phenomena.
I've never really understood the semantics of neo-, and perhaps because I'm all about semantic reflection at the moment, I gave it some deductive thought. Perhaps many of you are well comfortable with this little prefix and its denotations and connotations. It seems to be used by social commentators a lot, e.g. neo-christians, neo-conservatives etc. I've always had the impression that those particular examples referred to particularly fundamentalist versions of christianity or conservatism.
It seemed to me, from Haugh's use of the term, that it refers to a 'back-to-basics' approach or philosophy, perhaps after some considerable 'straying' over time from the origins of a movement or philosophy.
So I looked it up in the OED, and lo and behold, neo- is used when:
Forming compounds referring to a new, revived, or modified form of some doctrine, belief, practice, language, artistic style, etc., or designating those who advocate, adopt, or use it.
I think the use of revived is key in this definition, and are evident in some of the examples given in the entry:
The extracts we have given serve to show the dogmatic assertiveness of the Neo-Buddhist philosophy.
And just to tie it all back in to liguistics, this derived form was also listed:
neography n. a new system or method of writing or spelling.
Subscribe to:
Post Comments (Atom)
7 comments:
Bulanjdjan Maïa said:
The classical definition of inference links it, I'd say, to the derivation of a general principle on the basis of observation of empirical ground (repetition of a phenomena). Hence, more than reasonning, it is empirical observation that is deemed weak.
But that's only a philosophical definition. I may know all that by heart and still wonder, by the way, how do we use these words in "non philosophical English"?
Bulanjdjan Maïa said:
But by the way, did you need that for your classes??
Bulanjdjan Maïa said:
Thinking about it, implication is also used in logics as the name of the rule "if p then q".
Then it would make imply synonymous to "entail".
Hmm, I'm not sure I understand what you mean by 'it is an empirical observation deemed weak'.
Repetition of a phenomena is understood as 'convention', and conventional phenomena are usually understood as 'encoded' (hence conventional implicature ~ entailment). Encoded meaning is usually understood as semantic meaning, while pragmatic meaning is understood as derived from the particular use in context (conversational implicature).
Of course pragmatic meaning can become conventionalised - especially when the context itself is regularly repeated. However, the difference between semantic meaning/conventional implicature and pragmatic meaning/conversational implicature, is that the inferred proposition of the latter can be 'defeased' (is this term used in philosophy/logic more generally?), or cancelled.
This has been brought on by another linguist questioning my use of terminology, and my resultant investigation to address that point.
Bulanjdjan Maïa said:
Well, I'm relying on what have become more or less classical philosophical definitions (I'say) of Humean inspiration.
But you seem to be refering specifically to Grice, which as I understand provides the basis for the understanding of those terms in the context of semantics.
Hence the discrepancies, but I thing it's interesting to see what those definitions have in common!
Bulanjdjan Maïa said:
Actually you're right, it's mainly induction, considered a kinf of inference, that relies on observation, not every kind on inference. I should be careful when posting while having breakfast before going to work in the morning! Anyway, I think the previous comments remains true.
I'm enjoining the comments. Indeed, 'infer', 'implicate', 'imply', 'entail', 'presuppose', 'suggest, 'hint', etc. etc. etc. -- can all be pretty subtle. Feel free to drop by the Griceclub.blogspot and share a comment or two! I too was intrigued by the neo-Griceans; and have started to use "palaeo-Griceans" to refer to stuff like Aristotle! (vide: Horn, "Greek Grice: conversational rules in the history of logic", etc.). Cheers.
Post a Comment