PLIN0020

Advanced

Semantic

Theory

## Linguistic data - Data collection is very important in science - In 'armchair linguistics', data are introspective judgments of speakers around you or your own ('informal experiments') - Linguistic experimentation - Observe people's <b>acceptability judgments</b> - Try to understand what's going on

## Syntactic data - Syntax is a theory about <b>grammaticality judgements</b> - Grammaticality judgments cannot be directly observed. They need to be inferred from acceptability judgements. - You can't ask 'Is this grammatical?' with a technical term 'grammatical' to a naive informant!!

Ex. "What did Ben ask whether I ate?" - How can we know if this grammatical or not, if all we can observe is acceptability? - In a formal experiment, you need a good baseline condition, and also many variants of this example. - If you are checking your own intuitions, you might do all this implicitly, but the logic should be the same. - Fieldwork is somewhere in between. You need to be more careful with data.

## Semantic data - Truth-value judgments - Is a sentence true in a given situation? - Does one sentence follow from another? - Felicity judgments - Is a sentence appropriate/usable in a given situation? - Any other 'expermental tasks'

## Never ask ## "What does this mean?"!!!

## 🙅"What does this mean?"🙅 - Because native speakers often don't 'know' - They might be able to paraphrase content words, but there are tricky cases, e.g. 'take' - There might be subtleties you might miss, e.g. 'will' vs. 'is going to' vs. progressive - Speakers are much worse at funtion words, e.g. 'the', 'any', 'whoever'

## Reliable data in semantics * Truth-value judgments * Inference judgments * Felicity judgments - You may start with translations, but take them with a grain of salt - Always check them with acceptability judgments with discourse contexts

"[...] a semanticist does not ask a native speaker to conduct semantic analysis. This includes generally avoiding questions such as ‘what does this word mean?’, ‘when is it okay to use this word?’ and ‘does this sentence have two meanings?’ Instead, we mainly proceed by requesting judgments on the acceptability of sentences in particular discourse contexts." (Lisa Matthewson 2011:270)

## Ex. Uyghur attitude reports First present the context, and then the sentence (Matthewson 2011). Ex. "Ahmet Aygülni yaxshi körimen didi"<br />(Ahmet said that I like Aygul) - Context 1: Ahmet says "I like Aygül!" - Context 2: Ahmet says "Yasu likes Aygül!" vs. "Ahmet Aygülni yaxshi körigenlikimni didi"

## Ex. Universal quantifiers in English 1. Every dog is brown 2. All (the) dogs are brown 3. Each dog is brown Want to test: - Felicity judgments - Different nouns - Different predicates

## Ex. Biased questions in English 1. "Is Alice left-handed?" 2. "Isn't Alice right-handed?" 3. "Alice is left-handed?" 4. "Alice is left-handed, isn't she?" Factors for making contexts - I'm looking for L/R-handed people. - I thought she was L/R-handed - Contextual evidence that Alice is L/R-handed