SEMANTIC MAPS presented by Daniel Boucník “I’ll give a tutorial “How to make semantic maps” on one specific example - indefinite pronouns. We will go away from phrases and look at morphemes and their meaning.” Grammatical items are multifunctional Semantic maps help us: -define functions of a grammatical item -reach universal conceptual space -compare different languages easily Why are semantic maps useful? “Before telling WHAT S.M. are, I tell you WHY they are.” Multifunctionality of grammar (1) English preposition on a. a cup on the table b. a fly on the wall c. keys on a hook d. leaves on a tree ❏Monosemy? ❏Polysemy? ❏Homonymy? CALL TO ACTION: “Is this an example of monosemy/polysemy/homonymy?” - No need to choose! … uses = contextual meanings … senses = conventional meaning … FUNCTIONS = uses & senses together BTW: I am using Haspelmath’s numbering for easier cross-referencing with the paper. Zoom in: Functions Functions: (support/vertical) (support/horizontal) (attachment/non-part) (attachment/part) ...How do we tell? (1) English preposition on a. a cup on the table b. a fly on the wall c. keys on a hook d. leaves on a tree “= different functions of a gram (e.g. preposition to)” Monosemy = Is there only one word with a very broad meaning? Polysemy = Is there only one word with multiple different meanings? Homonymy = Is it actually 4 different words! The answer: SEMANTIC MAPS Haspelmath explains different modes of explanation, but arrives at SEMANTIC MAPS. SEMANTIC MAPS How to describe functions of words without polysemy/homonymy And finally… How the atoms of multifunctional grams are related in the conceptual space What is the terminology of semantic maps? Does everyone understand these terms? ❏Grams = grammatical morphemes ❏Functions = the senses and uses of a gram ❏Multifunctionality = the many functions of one gram ❏Conceptual space = the result of comparing functions across Ls Using “grams” is unnecessary, but economical. Ordering: G -> Fiction -> Multi -> Concept. space How S.M. work: ❏Find a gram which is multifunctional ❏Define distinct functions (at least 1) ❏Cross-linguistic comparison Example: Indefinite pronouns ❏Some-series ❏someone, something, somewhere, … ❏Any-series ❏anyone, anything, anywhere, … ❏No-series ❏no one, nothing, nowhere, … Grams = some/any/no How S.M. work: ❏Define distinct functions ❏Cross-linguistic comparison 1st part complete How do some-indefinites and any-indefinites differ? 1.Assertiveness (English) 2.Specificity (Russian X English) 3.Specific known/specific unknown (German X English) 4.… … Start with 1. Assertiveness - “They distinguish a degree of assertiveness.” Extend with 2. Specificity - “ Further extend 3. Specific known X specific unknown And that is not all... Function No.1 – Assertiveness (English) Some-indefinite – assertive contexts (5a-b) Any-indefinites – non-assertive contexts (6a-c) (5) a. Yesterday Mariamu met someone (/*anyone) from Botswana. b. At the DGfS conference I always meet someone (/*anyone) I know. (6) a. Has anything happened while I was away? (a question) b. If I can help you in any way, please tell me. (a conditional protasis) c. I didn’t notice anything suspicious. (a negative sentence) In declarative sentences truth is asserted. In questions, conditional protases and negative sentences truth is not asserted. Many functions still unexplored Any-indefinites – not (normally) possible in imperatives (7a) Some-indefinites – possible in questions and conditionals (7b-c) (7) a. Please buy something (/??anything) for our son when you go to town. b. Has something happened while I was away? c. If I can help you in some way, please tell me. We can see that some-indefinites and any-indefinites differ in assertiveness, but a lot of their functions are still unpredicted / unexplored / hidden. How S.M. work: ❏Cross-linguistic comparison Cross-linguistic comparison (En X Ru) Russian grams: ●-to ○Kto-to “someone” ○Čto-to “something” ○Gde-to “somewhere” ●-nibud ○Kto-nibud “anyone” ○Čto-nibud “anything” ○Gde-nibud “anywhere” Let’s look at other Ls! =D “English and Russian have two different grams in indefinite pronouns. The meanings of these grams overlap, but are not the same. The labels for the functions of these grams (assertiveness and specificity) are not sufficient to describe the full range of functions of these types of indefinites. BTW:(I’m using latin letters.) New functions to the map … Russian distinguishes something that is ambiguous in English (9) a. On xočet ženitsja na kom-to iz Botsvany. “He wants to marry someone [specific] from Botswana.” b. On xočet ženitsja na kom-nibud iz Botsvany. “He wants to marry someone [non-specific] from Botswana.” -nibud-indefinites occur when non-specific reference is intended (= Speaker does not have a particular referent in mind.) How S.M. work: ...What is next? 1st part complete… What’s next? “ Repeat for multiple Ls … in this case, 40 languages For more info see Haspelmath, M. (1997) Indefinite pronouns. Oxford: Oxford University Press) “I’ll focus on these terms “specific known” X “specific unknown.” Irrealis non-specific - (9b), (10) = something like “other” non-specifics Specific known, indirect negation, comparative = some Ls have indefinite pronouns that differs from each other in these functions One more L – German: (12) a. Mein Handy ist weg, (irgend)jemand muss es gestohlen haben. “My cell phone is gone, someone must have stolen it.” b. Jemand /*irgendjemand hat angerufen – rate mal wer. “Someone called — guess who.” Eng: “someone” Ger: Jemand or Irgendjemand? Specific known or specific unknown? … Irgendjemand cannot be used when the referent’s identity is known to the speaker. One more example: Both German words would be translated as “someone” but… A SEMANTIC MAP A SEMANTIC MAP – the English grams Some-indefinites – assertive contexts (5a-b) Any-indefinites – non-assertive contexts (6a-c) Remember? A SEMANTIC MAP – the Russian grams -nibud-indefinites – non-specific reference -to-indefinites – specific reference What is this good for? in Leipzig VS in February ❏Two uses of one morpheme for different contexts? ❏Two clearly distinct different morphemes? Generality fallacy = The generalization a linguist has made is the speaker’s generalization. We don’t know if in has two different meanings, really. Polysemy fallacy = A linguist makes many distinctions, and they say these happen in the mind of a speaker. We don’t know if , really. Semantic maps allow us to explore the cross-linguistic reality: some Ls use different prepositions for different functions What is this good for? We do not have to decide! SEMANTIC MAPS allow us to explore a cross-linguistic reality while avoiding hasty conclusions Haspelmath, M. (2003) “The Geometry of Grammatical Meaning: Semantic Maps and Cross-Linguistic Comparison” in: Tomasello, Michael (ed.) 2003. The New Psychology of Language, vol.2. Mahwah, NJ: Erlbaum, pp. 211-242