What makes languages similar
- Genetic relations
- Contact: borrowing
- Universals
- Coincidence
Similarity of form
- Cognates
- False friends
- embarrassed, Spanish embarazada 'pregnant'
- sympathetic, Spanish simpático, etc. 'nice'
- igloo, Inuktitut ᐃᒡᓗ iglu 'house'
- Japanese ベースアップ beesu appu (base up) 'raise in the base salary'
Lexical differences
- One-to-many lexical relationships
- Semantic dimensions; kinship terms
- Inclusion: abstractness, generalization
- parrot, squash
- blue, green; Russian синий sinij, голубой goluboj
- carry, eat in Mayan languages; wear in Japanese, Korean
- water, Japanese 水 mizu, 湯 yu
- Complex lexical semantic overlap
- Spanish for take
Grammatical-lexical differences
- Aspectual differences
- know, Spanish conocer
- put on, wear, Japanese 着る kiru; die, dead, Japanese 死ぬ shinu
- Progressive in English vs. French or German
- Differences in derivational morphology
- Valency-shifting morphology in Swahili
- Decreasing number of arguments
- Passive
Juma amefungua mlango. 'Juma has opened the door.'
Mlango ulifunguliwa. 'The door was opened.'
- Stative
Mlango umefunguka. 'The door was opened.'
Mlango unafunguka. 'The door is openable.'
- Reciprocal
Juma anapenda Fatuma. 'Juma loves Fatuma.'
Juma na Fatuma wanapendana. 'Juma and Fatuma love one another.'
- Increasing number of arguments
- Causative
Juma alihama. 'Juma moved.'
Serikali ilimhamisha Juma. 'The government moved (resettled, expelled) Juma.'
- Applicative (prepositional)
Juma aliwafungulia wageni mlango. 'Juma opened the door for the guests.'
- Combinations of valency-shifting morphemes
- Fundi alitengeneza gari. 'The mechanic repaired the car.'
- Gari lilitengenezwa (na fundi). 'The car was repaired (by the mechanic).
- Fundi alimtengenezea Juma gari. 'The mechanic repaired the car for Juma.'
- Juma alitengenezewa gari (na fundi). 'Juma was repaired-for the car (by the mechanic).
- Differences in what's optional, obligatory, unmarked
- Number
- Quasi-semantic noun or verb categories
- Syntactic differences
- like, Spanish gustar, German gefallen
- hungry, thirsty, tired, bored: adjectives vs. verbs
I'm tired.
Amharic ደክሞኛል dǝkmoññal
lit. 'it has tired me'
Pragmatic differences
- Personal pronouns
I, me, Japanese わたくし watakushi, わたし watashi, あたし atashi, 僕 boku, おれ ore
- Situation-specific formulas
Gaps
- Lexical gaps
- ተግደረደረ tǝgdǝrǝddǝre 'pretend to have had enough of something when offered more'
- Sound symbolism, expressives
Lexical differences: origins and representation
- Where do the categories associated with everyday words come from?
- They are largely universal because they originate in non-linguistic cognition; word learning
is mapping words to already existing concepts.
- Lexical categories are free to vary (within limits) according to cultural, historical, and environmental differences; learning words involves working out language-specific principles of categorization.
-
Approaches to representation
- Intensional: feature-based compositional representations
- Example: break: x cause
[y become broken]
- Primitive predicates, e.g., cause
- Constants, e.g., broken
- Extensional: representation in terms of the things (events, etc.) referred to
The cut/break project (Bowerman, Majid, Senft, etc.)
- "Cutting and breaking": separation in the material integrity of objects
- Persistent errors in child language indicate that children are working out the categories themselves
- Method
- An extensional approach to lexical semantics
- Elicitation of data: 61 videoclips depicting separations of various kinds
- Event description collected for 28 languages from 18 families
- Problems figuring the unit of analysis
- Easy cases: single monomorphemic verb
- Compound verbs in Mandarin: bai1-duan4 'snap-be.broken.crosswise (of a long thing)', jian3-kai1 'cut.with.
scissors-be.open'; both verbs included
- Germanic verbs plus particle or other complement: break apart, smash to bits; "satellites omitted"
- Differences in number of lexical categories: 3 (Yélî Dnye), 50+ (Tzeltal)
- Perceived similarity of events: extent to which speakers of a given language used the same predicate to describe videoclips
- 28 × 61 = 1830 language/videoclip pairs subjected to correspondence analysis, multivariate statistical technique which extracts dimensions in order of importance
- First few dimensions separated simple separation verbs (opening, taking apart, peeling, etc.) from core cut/break verbs involving destruction
- Second correspondence analysis on core cut/break verbs
- Dimension 1: predictability of locus of separation; extremes distinguished in all sample languages
- Dimension 2: distinguishes hand action on 2-dimensional flexible object (tearing) from other actions
- Dimension 3: distinguishes smashing a rigid object with a sharp blow from snapping a long object into two pieces;
languages differ in terms of whether they respect this distinction, fail to, or give a choice
Summary of differences
- One-to-one, but sometimes word-to-multiword expression or MWE-to-MWE
- regret, arrepentirse de; I regretted coming, me arrepentí de haber venido
- put on, ponerse; he put on the coat (put the coat on), se puso el abrigo
- One-to-many, often with a "default"
- carry; Tseltal: chaw (in hand), chup (in shawl), k'ech (in arms, on shoulders),
kajan (on top), kuch, lep (loads), lutz' (hidden in clothes), pach (upright),
pet (in arms)
- French passé composé, English simple past and present perfect;
je les ai vu, I saw/have seen them
- Many-to-many
- take, carry, wear, use, drink; llevar, tomar, usar, beber
- Gaps
- concuñado/-a, 'brother-in-law of wife, sister-in-law of husband
