SEMANTIC PRIMITIVES.

A post at Taccuino di traduzione about the linguist Anna Wierzbicka and her “semantic primitives” led me to the natural semantic metalanguage (NSM) homepage, which provides a basic introduction to the theory:

The approach is based on evidence that there is a small core of basic, universal meanings, known as semantic primes, which can be found as words or other linguistic expressions in all languages. This common core of meaning can be used as a tool for linguistic and cultural analysis: to explicate complex and culture-specific words and grammatical constructions, and to articulate culture-specific values and attitudes (cultural scripts), in terms which are maximally clear and translatable. The theory also provides a semantic foundation for universal grammar and for linguistic typology. It has applications in intercultural communication, lexicography (dictionary making), language teaching, the study of child language acquisition, legal semantics, and other areas.


As someone trained in the more philological branches of linguistics, immersed in the maddeningly diverse variables of language in all their attested detail, I’m naturally suspicious of all notions of “universal grammar” and other universalities, but I do find the idea of “semantic primitives” (or “primes”) appealing, and I’m curious to see to what extent they can be found across the spectrum of languages; you can read about them, and see the current list (expanded from an original 14 to around 60) here. Some qualifications:

When we say that a semantic prime ought to be a lexical universal, the term “lexical” is being used in a broad sense. A good exponent of a primitive meaning may be a phraseme or a bound morpheme, just so long as it expresses the requisite meaning. For example, in English the meaning A LONG TIME is expressed by a phraseme, though in many languages the same meaning is conveyed by single word. In many Australian languages the primitive BECAUSE is expressed by a suffix.

Even when semantic primes take the form of single words, there is no need for them to be morphologically simple. For example, in English the words SOMEONE and INSIDE are morphologically complex, but their meanings are not composed from the meanings of the morphological “bits” in question. That is, in meaning SOMEONE does not equal “some + one” and INSIDE does not equal “in + side”. In meaning terms, SOMEONE and INSIDE are indivisible.

Semantic primes can also have variant forms (allolexes or allomorphs); for example, in English the word ‘thing’ functions as an allolex of SOMETHING when it is combined with a determiner or quantifier (i.e. this something = this thing, one something = one thing).

Exponents of semantic primes may have different morphosyntactic characteristics, and hence belong to different “parts of speech”, in different languages, without this necessarily disturbing their essential combinatorial properties.

And I find this quote from Goddard and Wierzbicka 1995 (Keywords, culture and cognition. Philosophica 55(1), 37-67) a useful way to think about the Sapir-Whorf hypothesis (I have changed “j” to “y” in the Russian quote):

Inye veshchi na inom yazyke ne myslyatsya ‘there are some things which cannot be thought in another language’, wrote the poet Marina Tsvetaeva… In a theoretical sense, this statement may be somewhat of an exaggeration, if, as the NSM theory contends, any culture-specific concept can be decomposed into a translatable configuration of semantic primes. … But in an important sense, Tsvetaeva’s statement remains true, because in practice it is impossible to formulate and manipulate thoughts of any sophistication without resort to the kind of conceptual ‘chunking’ enabled by the use of complex lexical items. Thoughts related to [Russian] dusha, for example, can be formulated in English only with great difficulty and at the cost of cognitive fluency, whereas in Russian they can be formulated more or less effortlessly.

I think it’s precisely the degree of difficulty, rather than the possibility, of formulation that lies at the heart of whatever truth there is in the hypothesis.

Comments

  1. The list seems heavily skewed toward representational language vs. “emotive” or interpersonal, language — please, thank you, I love you, I hate you, do [this] (command), etc.
    This is characteristic of AI people; they write the impersonal objectivity required to do good AI work into the intelligence they’re producing. Likewise it’s characteristic of much modern philosophy since Locke/ Descartes, especially of the possitivist sort. You still frequently have people who divide mental behavior into “rational” (narrowly defined) and “irrational” — with irrational meaning defective or harmful.
    Historically and functionally I think that this is a descriptive error. One of the evolutionary advantages of language was the enabling of cooperative action within groups defined in terms of affection, relationship, and heierarchy.

  2. I quite agree with you, and I’m surprised the bias has lasted this long — another aspect of the “two cultures” divide, I guess.

  3. this sort of thing has long been a
    staple of conlangers–Lojban & Vorlin
    having made significant compilations–
    but when i actually look at such a list,
    i find myself coming up with exceptions
    almost at once. KNOW–many Romance
    languages distinguish know-from-experience
    from know-by-being-told: & i can’t see
    that this is any less basic than DOG vs.
    CAT. but it is interesting to try it, all
    the same…

  4. This reminds me of Swadesh-type basic-word lists. Not very useful, perhaps, in strictly realistic terms. But they are beautiful lists, though:
    I
    thou
    we
    this
    that
    who
    what
    not
    all
    many
    one
    two
    big
    long
    small
    woman
    man
    person
    bird
    dog
    louse
    tree
    seed
    leaf
    root
    bark
    skin
    meat
    blood
    bone
    fat (n.)
    fire
    egg
    horn
    tail
    feather
    fish
    hair
    head
    ear
    eye
    fingernail
    nose
    mouth
    tooth
    tongue
    foot
    knee
    hand
    belly
    neck
    breast
    heart
    liver
    drink
    eat
    bite
    see
    hear
    know
    sleep
    die
    kill
    swim
    fly (v.)
    walk
    come
    lie
    sit
    stand
    give
    say
    sun
    moon
    star
    water
    rain
    stone
    sand
    earth
    cloud
    smoke
    ashes
    burn
    road
    mountain
    red
    green
    yellow
    white
    black
    night
    warm
    cold
    full
    new
    good
    round
    dry
    name

  5. I don’t know about the Swadesh list but some lists of this type are very useful, e.g. for missionaries or anthropoligists trying to puzzle out a totally unknown language. You can use it to organize the research and coinstruct a kind of pidgin.

  6. Expressions such as “love” are not excluded from the primatives list because they were not considered. I’d venture to say that most, if not all, languages have something more-or-less equivalent. Rather, they are not included because the words can be broken down into a “cultural script” which unambiguously describes their essential properties to both native speakers and foreigners. Even though some properties of the word “love” might be shared between many different cultures, other properties might need a surprisingly long explanation. For example, imagine a culture in which it is “loving” to kill someone rather than to let him live dishonored.

  7. Someone mentioned that some languages distinguish between know from experience and know from being told. But these English expressions show that neither of these concepts are necessarily “primitive” concepts (despite them being expressed in single words by some languages): it is possible to decompose those concepts into “KNOW” + by being told vs “KNOW” + by experience.
    In English you could phrase this as “I know this because someone said this to me” vs “I know this because this happened to me”, both of which are valid “explications” in the English natural semantic metalanguage. They are composed entirely of semantic primes: I, KNOW, THIS, BECAUSE, SOMEONE, SAY, for the first sentence, and the second sentence using the same semantic primes, except SAY is replaced by HAPPEN.
    The words in capitals are the English names for these semantic primes. It’s important to note that the concepts themselves are *not* defined by these labels, they are just useful for glossing. For example, the word “me” is an allolex of the word “I”, which are both representative of the same semantic prime. (The idea of allolex is a theoretical tool used in natural semantic metalanguage theory to rigourously define the idea of words being synonyms. Basically, two words are allolexes if they are used when saying “the same thing” in different grammatical contexts. For example, in the sentence “I read the book”, it would be ungrammatical to say “Me read the book”, and similarly, the sentence “Sam gave me a book”, it would be ungrammatical to say “Sam game I a book”, so they are in complementary distribution.
    The theoretical claim is that you could do something similar with the Romance words for know by being told versus know by experience.
    You could say “I know-by-experience this because this happened to me”, vs “I know-by-being-told because someone said this to me”. Assuming it would be incorrect to swap the words for “know-by-experience” and “know-by-being-told” in these sentences, that shows they are allolexes: one form of KNOW is used when combined with SAY, while the other is used when combined with HAPPEN. (And if it wouldn’t be incorrect to swap the words in those sentences, then that distinction isn’t actually obligatory in those languages, so it becomes a matter of different style rather than a matter of different meanings)
    I strongly recommend people to read Semantics: Primes and Universals by Anna Wierzbicka, it’s a really fascinating book! I only have access to the preview available on Google books, but even that bit is enough to start explaining these ideas. 🙂
    The idea is that these semantic primes are found in *all* languages; the fact that English doesn’t make this distinction shows that it’s *not* a semantic prime (and the same idea applies to distinctions that are not made in any other languages). Semantic primes look for concepts that are universal across all human languages.

  8. Here‘s the current Natural Semantic Metalanguage (NSM) page.

  9. David Eddyshaw says:

    Hausa ji = understand, hear, feel, taste, smell.

    For a start …

  10. David Marjanović says:

    French sentir = feel, smell.

  11. Russian слышать ‘hear; feel, sense; smell.’

  12. David Eddyshaw says:

    I notice that “smell” actually doesn’t figure among the list of supposed primitives: it seems reasonable to take it as “sense by-nose.” But then I can’t see any very profound reason not to analyse “see” likewise as non-prime: “sense by-eye.”

    “Hear/smell” often go together in West Africa: Kusaal has wʋm, “hear; smell; understand (speech.)”

    There is a completely homophonous (including tone) wʋm “afflict, bother”, but I think that’s probably accidental; other Western Oti-Volta languages have homophonous cognate pairs of both verbs, though, so I suppose it’s conceivable that there is after all some semantic connection which I can’t see from the standpoint of SAE. In Hausa you ji “hear” blows to the body (as in the proverb jiki ya fi kunne ji “the body hears better than the ear”, i.e. if you don’t listen to advice you’ll learn the hard way.) So conceivably wʋm “afflict” is an old causative formation from wʋm “hear” (that’s morphologically possible, at any rate: the causative suffix –l would be assimilated to the preceding m.)

  13. David Eddyshaw says:

    I don’t know anything about this, but (as usual) feel in no wise inhibited from commenting on that account.

    This seems to me to have the sulphurous whiff of logical atomism, the approach which St Ludwig expressed so attractively in his Tractatus and spent the rest of his career utterly demolishing after he realised it was bollocks following his spell as a primary school teacher (this should probably be compulsory for all philosophers.)

    More specifically: these “primes” aren’t coherent cross-linguistically: semantic fields don’t match between languages, and this is not only true of conceptually complex things like “boredom” but equally (or more) true of these alleged atoms like “feel”, as we’ve been saying above,

    The notion of “molecules” is even more shaky. There is no single coherent category “child” or “bird” (for example) which works cross-linguistically: to suppose there is is to ignore all the work on the subject of categorisation that has been going on since Wittgenstein.

    The question of how these supposed atoms and molecules can be combined both syntactically and conceptually must not be hand-waved away: in no real human language is it possible to analyse such processes in such a way that complexes are simply the sums of their parts: the very processes of composition themselves are frequently irregular, unpredictable and idiosyncratic: in other words, they introduce more “primitives” by stealth.

    The idea that you can identify primitives by looking for highest common factors cross-linguistically strikes me as fundamentally circular: you factor out everything that actually matters in the process.

    [On the other hand, John Cowan seems a great fan of this stuff, and I am always (rightly) uneasy to find myself on the other side of an argument from him …]

  14. David Eddyshaw says:

    The danger sign is when you end up labelling words like (for example) “do” as primitive. The thing about English “do” is that it has little or no semantic content at all; its importance lies in the central part it plays in syntax. The idea that it plays a central role in meaning is surely so bizarre as to be a reductio ad absurdum.

    Et sic de similibus.

  15. in meaning SOMEONE does not equal “some + one”

    It doesn’t?

  16. I get more sceptical of this sort of thing as I get older. In fact, I’m getting more and more sceptical about received ‘grammar’, too.

    English has two articles: ‘a’ and ‘the’. They are articles because at some stage someone decided they are ‘articles’, in parallel with other European languages. So they must be articles.

    As a matter of semantics, however, the demonstrative ‘this’ can be used in much the same meaning as ‘a’ in colloquial speech. (Why does this tend to be ignored? Because of prescriptivism and a bias towards written English, maybe?)

    This guy came up to me and started arguing about semantic primitives.

    In this case, ‘this’ functions in the same way as ‘a’. It refers to a person unknown to the listener but known to the speaker. While it is INDEFINITE, however, it differs from ‘a’ in having only a SPECIFIC meaning — one particular guy.

    ‘A’ can be either specific or non-specific. The classic example is:

    Mary wants to marry a Norwegian.

    where ‘a Norwegian’ could be non-specific (any Norwegian will do) or specific (there is a specific Norwegian whose identity is, however, unknown to the listener at the time of speaking). If you use ‘this’ the meaning can only be specific.

    Mary wants to marry this Norwegian.

    So colloquial English draws semantic distinctions that standard written English does not.

    ‘Some’ can also be used in the same position. I will leave out a consideration of the semantics of ‘some’ here (I’m sure semantic primitives theory can come up with something), but how does it capture the dismissive tone of ‘some’ in this sentence?

    Mary wants to marry some Norwegian.

    Semantic primitives sounds like a useful tool but I would suggest it has its limits.

    Did you know that women on the contraceptive pill are supposed to be 10% less sensitive to expressions of subtle emotion like contempt than women who are not? (Has to be true, I saw a video on the Internet.) Well, perhaps linguists who try to stuff language into theories are (at least) 10% more insensitive to capturing the nuances of real language than those who don’t.

  17. Stu Clayton says:

    I don’t know anything about this, but (as usual) feel in no wise inhibited from commenting on that account.

    That’s not hard or unusual. si exemplvm reqviris, circvmspice. The inhibitions advene in the replication, when those who do know something about it come down on you like a ton of bricks.

    I agree 100 pro about the sulphurous whiff of logical atomism.

    as in the proverb jiki ya fi kunne ji “the body hears better than the ear”, i.e. if you don’t listen to advice you’ll learn the hard way

    Wer nicht hören will, muß fühlen. Spare the rod, spoil the child.

  18. Stu Clayton says:

    Well, perhaps linguists who try to stuff language into theories are (at least) 10% more insensitive to capturing the nuances of real language than those who don’t.

    I conclude that you are not on the pill. The blue pill, that is.

    No systematic theory can cover everything. Whatever a “theory” is, it cannot be in 1-1 correspondance with reality, since that would leave no wriggle room for thoughts, doubts and objections. It would be merely a heaving mass of interlocking causalities, exactly like what it proposes to describe “objectively”.

    Also, since a theory is part of reality, which part of it could be in 1-1 corespondance with itself ? Grammar is not a subfield of transfinite arithmetic. Not the last time I checked, at any rate. I try to monitor all the latest developments.

  19. ‘a’ can be either specific or non-specific.
    Mary wants to marry a Norwegian.

    But, but, the ambiguity there is nothing to do with ambiguity of words; and indeed is an ambiguity that the logical atomists/Positivists have characerised thoroughly. Bathrobe is about 120 years behind the times. There’s two distinct scopes of the quantifier hidden inside the indefinite article ‘a’.

    There is some Norwegian x such that Mary wants to marry x.

    For any x that Mary marries, she wants x to be Norwegian.

    (There’s also ambiguous scoping of what Mary wants, but let that pass.)

    I guess a Generativist might say there are two different deep structures that get realised in the same surface utterance. Just as in ‘Mary is easy to please’/’Mary is eager to please’, the difference in meaning is not merely down to choice of vocab.

    I agree that if somebody were claiming all meaning is down to semantic primitives rather than syntax/logical structure, that would be a travesty. Such a claim was made by Katz & Fodor’s semantics, with blessings from Chomsky at the time. (They’ve all retreated shame-facedly since.)

  20. PlasticPaddy says:

    I thought the use of the atoms was to build a sort of cross-compiler for machine translation or to support “first principle” work in anthropology. You can always say that some languages have “redundant” or “defective” structure, i.e., English instantiates “GO BY FOOT” as a single word whereas German and French do not, or you could say German has lost “GO” (instantiated redundantly with “GO BY FOOT”).

  21. Stu Clayton says:

    But, but, the ambiguity there is nothing to do with ambiguity of words; and indeed is an ambiguity that the logical atomists/Positivists have characerised thoroughly. Bathrobe is about 120 years behind the times.

    But, but, Bathrobe did not write about “ambiguity”. He doesn’t even use the word “word”. This is what he wrote, for example, after having demonstrated it: “So colloquial English draws semantic distinctions that standard written English does not.”

    You have jumped the gun, but I suppose that’s OK since no one else is competing in this dash to find fault.

  22. colloquial English draws semantic distinctions that standard written English does not.

    Cross out “standard written English”, insert any language you like. That claim is looking a lot like the ‘no word for X’ nonsense.

    Of course no language (colloquial or otherwise) draws distinctions un-drawable in any other language. It might be that in the other language it takes a longer phrase/restructuring the whole utterance. That’s in effect what I did with the logic-style quantification. It might be that to draw out the dismissive tone of “some Norwegian” would take a lot rephrasing, to the extent it falls flat like explaining a joke. But to claim a distinction can’t be expressed in written English and then go on to express it in written English is … words fail me.

    Bathrobe’s “‘a’ can be either …” that I quoted is talking about a word. Bathrobe goes on to use that word in a sentence that has two readings (at least — actually several others not considered). That Bathrobe did not use the word “word” nor the word “ambiguity” is neither here nor there.

  23. David Marjanović says:

    in meaning SOMEONE does not equal “some + one”

    It doesn’t?

    I’d say it equals some + person, or some + one + person depending on the language.

    English instantiates “GO BY FOOT” as a single word whereas German and French do not

    Depends on the kind of German; mine uses gehen for that purpose by default (though zu Fuß gehen is often specified to stress the contrast to fahren), over here laufen is common (which for me means “run” instead).

    In French, how about marcher? It’s not used very often in this literal sense, but still.

    or you could say German has lost “GO”

    Mine has. I have no single word for “locomote”. Gehen means by foot, fahren means on wheels, by sleigh/sled/skis, by ship, to hell, or up into heaven (in den Himmel auffahren, not to be confused with Auffahrunfall, the kind of accident where a car crashes into a slower car, “going up” on it)…

  24. This seems to me to have the sulphurous whiff of logical atomism

    To me as well.

    Mine has. I have no single word for “locomote”.

    You have become a Slav!

  25. John Cowan says:

    since a theory is part of reality, which part of it could be in 1-1 corespondance with itself

    Consider a portrait of a painter. In the painting, the body of the painter isn’t really flesh, it’s just paint. Likewise, the palette isn’t really wood, it’s just paint. And by the same token, the paint on the palette isn’t reallly paint, it’s just — No, wait, it really is paint!

    In sympathetic magic, the best symbol for a sharp knife is probably a sharp knife.

    Russian слышать

    Unlike sentir, this is plainly a cognate of listen, thanks to the magic of s-mobile, and thus semantically like Hausa ji.

    The notion of “molecules” is even more shaky.

    Molecules have no theoretical significance: they are simply compact ways of writing a particular explication, like acronyms, except that acronyms can take on meanings of their own: not every instrument for light amplification by stimulated emission of radiation is what non-technical anglophones would call a laser. As such, their English meanings are even more misleading than those of the primes.

    [DO] as primitive

    If I were Anna Weetabix, I would have consulted a colleague at ANU and given all the primes names in Sakao instead of English. In that way, if the words were artificially narrowed or widened in meaning, only a few thousand native speakers and Jacques Guy would complain.

    And given how holophrastic and polysynthetic Sakao is, it would certainly silence the objection “Why, some of your so-called primes aren’t even single words!” I mean, would you expect œsɨŋœɣ ‘my mouth’, œsɨŋœm ‘thy mouth’, ɔsɨŋɔn ‘his/her/its mouth’, œsœŋ-PN ‘PN’s mouth’ in an Oceanic language? Or a sentence mɔ-sɔn-nɛs-hɔβ-r-ɨn a-ða ɛ-ðɛ ‘s/he-shoots-fish-follows-CONT-TRANS ART-bow ART-sea’? (Sakao’s close relartive Tolomako, spoken in the same place, has na tsiɣo-ku, na tsiɣomu, na tsiɣo-na, tsiɣo-PN respectively; Sakao and Tolomako are more or less the Spoken French and Written Sardinian of Vanuatuan languages, except even further apart. )

    no language (colloquial or otherwise) draws distinctions un-drawable in any other language

    It might have to be explained from scratch. There is no French way to say bobbing for apples, because as Marie-Lucie (eheu fugaces!) said, no French person would engage in such a silly pastime. But NSM doesn’t care how many words it takes to express a semantic prime in a particular language (see Sakao above).

    So colloquial English draws semantic distinctions that standard written English does not.

    High-register Standard English certainly does draw this distinction: “Mary wants to marry a certain Norwegian” is impeccable. In Lojban, where the articles historically marked specificity (more recently the formerly non-specific article has been used without reference to specificity) and definiteness is a strictly optional particle, a certain X is a very common bit of translationese.

  26. My point wasn’t that phenomena such as scope, negation, etc. in natural languages can’t be described through formal logic. They can and logicians have done it. But i if you care to check, you will find that ‘specificity’ and ‘definiteness’ are established terms in linguistics, despite the existence of formal logic to do the job. Specificity_(linguistics)

    My point was that linguistics and grammar seem to have turned into grand traditions with a life of their own. Do we even need the category of ‘article’ in English grammar? Isn’t it only inertia (and perhaps interlinguistic usefulness) that keeps the category alive?

    Labov (I think it was) who described the tense / aspect system of Black English. Until he turned a linguist’s eye on it, it was denigrated as ‘substandard’ or worse. The appropriate response to the fact that Black English draws distinctions in a different way from standard English is not to say that “Standard English can draw those too, if it wants”. It is to give recognition to the existence of different systems — even if they can be described in terms of formal logic.

    I left the relationship of my comments to the issue of semantic primitives vague, but a similar logic can be applied to ‘know’. Sure, English can draw a distinction between ‘know through hearsay’ and ‘know through experience’. So what? Essentially, I am sceptical of any system that seems to take English as standard, just as I am sceptical of grammatical traditions that continue for traditions’ sake. Surely Sakao would be just as good.

  27. David Marjanović says:

    You have become a Slav!

    I’ve noted the similarity before, but I’ve also pointed out that, at least in Russian, you swim by ship.

    Russian слышать

    […] this is plainly a cognate of listen

    (Imagine a h- in front of listen, and it almost makes sense.)

  28. Yup. The OED says listen is from “Old Northumbrian lysna, *hlysna, corresponding to Middle High German lüsenen < Old Germanic type *hlusinôjan, < Germanic root *hlus– : see list n.1,” and s.v. list we get:

    Old Germanic *hlusti-z < Old Aryan *klusti-s (Sanskrit çrušti obedience), < root *klus– (:kleus– :klous-), Old Germanic *hlūs– (:hleus– :hlaus-), found also in the verbs Old English hlosnian, Old High German losên (Middle High German losen), Old High German lûstrên (modern German dialect laustern: compare German lüstern, Swedish lystra, Danish lystre to ‘answer’ to a name, ‘answer’ the helm), Middle High German lûschen (modern German lauschen), Middle High German lusemen, lüsenen, all meaning ‘to listen’; also, outside Germanic, in Old Church Slavonic slyšati to hear, sluχŭ hearing, Lithuanian klausà obedience, klausýti to hear, Avestan çraosānē to hear, Welsh clûst, Irish clúas (feminine), ear (< Old Celtic *kloustā).

    Those entries are from 1903, mind you.

  29. David Eddyshaw says:

    If Semantic Primes are supposed to exist independent of any particular actual language, I don’t see how that is even possible, unless you either believe in Mentalese, or you interpret the claim so weakly that you end up with a set of constraints which simply arise from our physical constitution, sensory equipment and so forth. It doesn’t seem to me to be a given even that human cognition necessarily divides experience into things and happenings, let alone anything more fine-grained.

    If the claim is the less ambitious one (as it seems to be) that as a matter of empirical observation, all human languages can in fact be factored in this way, that is the very claim which seems to me to be fatally prone to circularity. This seems to be the sort of theoretical framework you could fit any empirical observation whatever into by refining your primitives by progressively blurring their edges while emptying them of specific content. I would dearly like to know what sort of possible linguistic observations would be taken as refuting the theory in principle.

    And the work seems to follow the familiar pattern of working out your universalist claims starting with the usual-suspect tiny subset of highly atypical languages like English, Spanish, Japanese, Chinese, then looking for confirmation of the validity your approach in a few “exotic” languages, and discovering to your surprise and delight that they are, after all, really just like the languages you are familiar with in every way that matters.

    The fact is that investigating the semantics of the language of a very alien culture is extremely difficult, and full of traps for the unwary, even if you don’t come with the baggage of a pet theory that you already committed to, instead of keeping your mind open as to what you might discover.

    In point of fact It does seem to me unarguably true that all the human groups we’ve ever met do in fact share whole vast realms of complex concepts – which they cover with different, mutually incongruous linguistic patchwork quilts. Why does what we have in common in this way have to be forcibly allocated (in the teeth of the evidence) to language specifically? Why not culture? If we all have ways of saying “child” (say), why is that a linguistic fact rather than a cultural observation? (particularly as our various human terms even for this most basic concept do not as a matter of fact, neatly coincide in reference.)

    In fact, I wonder if this approach is another symptom of the abduction of linguistics in the course of the twentieth century from its true home in anthropology and its press-ganging into the realm of (individual) psychology.

    (I should say – hoping to stir JC in particular into action – that these are not rhetorical questions. The Semantic Primers are evidently neither ignorant nor stupid, and may very well have anticipated all these questions and even come up with convincing answers. With any luck JC will convince me and save me from the trouble of actually reading their books …)

  30. David Eddyshaw says:

    Molecules have no theoretical significance: they are simply compact ways of writing a particular explication

    Ah. I was thinking they were a sort of escape hatch from some of the more implausible consequences of the theory. But that actually makes the implausibility pretty insuperable.

    I frankly disbelieve that, say “bird” (one of the examples given of a “molecule”, and actually one the more conceptually straightforward) can in fact be decomposed into Linguistic Primes at all in any language. The assertion that in principle it could be seems quite groundless. It’s based on a Leibnizian conception of how words refer which Just Doesn’t Work.

    Even if there were an unambiguous definition, that is not what the word means; that is determined by how the word is used. Dictionary definitions are a pis aller: little summaries or essays to give the reader an idea of how the word is used. Otherwise the book would need to be quite unwieldy …

    The idea that the true meaning of a word can be discovered by semantic factoring is just a fancier version of the etymological fallacy. It can work to some extent for a small subset of the vocabulary: structured semantic fields like kinship terms, for example. Even then, real languages escape the straitjacket: Kusaal diem, for example, is the kinship term for “wife’s mother or father.” “Husband’s mother or father” is a different word, dayaam, and diem cannot be used in this sense. However, diem is used in polite address by individuals of either sex to address an unrelated person of opposite sex and similar or greater age to oneself. Factor that.

  31. In an impressionistic, roundabout way I am agreeing with David Eddyshaw.

    My comments above were directed at traditional grammar. The Cambridge Grammar of the English Language does, in fact, subsume the articles under the determiners, noting that they provide the most basic definition of definiteness and indefiniteness. It does not, however, appear to cover the usage of this that I mentioned above, which I regard as a fault. If you are ignoring such phenomena in your own language due to inertia, how can you approach others in an intelligent manner? (Or: how the hell can you have a grammar of English that ignores colloquial English?)

    Some generative grammarians extend the ‘determiner phrase’ to languages that don’t even have articles, presumably on the basis that ‘the articles are there but are realised as zero’.

    I wholeheartedly agree that universalist claims all too often start with “the usual-suspect tiny subset of highly atypical languages like English” etc.

    This is why I expressed doubt, perhaps tinged with cynicism, about old, at times rotting parts of the edifice that form the core foundations of our “knowledge”. The anthropological structuralists seem to have had a better approach — don’t take your own assumptions for granted when approaching totally different languages — than either traditional grammar or generative grammar.

  32. David Eddyshaw says:

    Actually, one of the very reasons I hero-worship the Cambridge grammar is that it makes no claim to being “Comprehensive” (although perhaps that’s because that title was already taken …)

    Instead, although it’s far and away the most impressive grammar of any language I’ve ever seen overall, it presents the facts in all their glorious messiness and is not shy of admitting that a lot more work needs to be done, or that there might be several perfectly good alternative ways of describing the material. And It shows its workings rather than presenting you with a flawless theoretical plan.

    When I (rarely) catch it out in error (or more often, incompleteness), I consequently feel a deep sense of achievement rather than a desire to ask for my money back.

    I’ve found it very useful in my trying to make sense of Kusaal grammar, not because Kusaal is particularly similar to English at all, but because CGEL again and again shows you the right sort of questions you should be asking yourself about how a language works, even when their own answers are unsatisfactory or avowedly incomplete.

  33. Bathrobe, as someone who is a native Russian speaker and who has no use for articles at all, I think you are approaching the matter from the (a?) wrong angle. Articles (in English) are little words that you have to stick in front of a noun, otherwise your sentence would be ungrammatical. All the semantic baggage that they bring is strictly secondary. Let’s take a simple Russian sentence “Картина висит на стене как-то косо”. Word for word English is “Picture is hanging on wall somewhat askew”. This misses 2 determiners out of 2. I cannot imagine any realistic situation where you can get away with not sticking something in front of either of the nouns. Yet in Russian nothing extra is required. Of course, you can stick this or many other words, but they will give the sentence extra meaning. The extra meaning given by a/the in this situation is the minimal possible info that English grammar allows to provide. The fact that in some other situation this provides as little information as a is not relevant because a fulfills this minimalist function routinely. I guess, this is the core of my argument, it doesn’t matter which function a word can play in any conceivable situation, it’s more about its core function.

  34. Articles (in English) are little words that you have to stick in front of a noun, otherwise your sentence would be ungrammatical.

    There is actually a difference in meaning depending on the choice of article. It’s not just a matter of providing a “grammatical filler”. Getting the article wrong can hamper comprehension — in fact, it might be better to leave them out if one can’t use them properly, at least at a fairly rudimentary level. What I mean here is, for instance, the use of ‘a’ when speaking of a person or thing the first time, and switching to ‘the’ thereafter. If an English learner continued to use ‘a’ right through, the effect could be bewildering.

    I agree that the Cambridge grammar is very good. My quibble is with the way that they deliberately ignore colloquial English.

  35. There is actually a difference in meaning depending on the choice of article.

    Sure. My point is that in a grammatically correct English the speaker has no choice but to supply this difference in meaning.

    in fact, it might be better to leave them out if one can’t use them properly

    That’s not what I was told. I was told that using the wrong article in front of a singular noun is less jarring than using none at all. But probably the tastes in ungrammaticality differ.

  36. David Eddyshaw says:

    in fact, it might be better to leave them out if one can’t use them properly

    Briefly introspecting (an unsafe guide), I think Bathrobe is right on this.

    An analogous case was when I lived in northern Ghana, where none of the local languages distinguishes male from female grammatically, including even the local instantiations of Hausa and quite a lot of otherwise perfectly good English (it’s hard to make distinctions consistently in a foreign language when they don’t exist at all in your own.)

    What regularly threw me even after years of it was not people who said “he” for “she” but those who said “she” for “he” (which, as I’ve said before, probably has uncomfortable implications regarding markedness in my own idiolect, if not in English in general.)

  37. David Marjanović says:

    I agree that the Cambridge grammar is very good. My quibble is with the way that they deliberately ignore colloquial English.

    Probably a conscious decision to limit the size of the book (and the work!). You’re not going to find a grammar of all of German, in part because it would take a lifetime to write.

    if not in English in general

    Probably Western culture in general. The only explanation I’ve ever found why, although women are now allowed to wear trousers, men still aren’t allowed to wear dresses or skirts (if they can’t get away with calling them kilts), is that women are now allowed to upvalue themselves, but dressing “like a woman” when you aren’t one is still subconsciously regarded as downvaluing yourself. Star Trek has predicted this state of affairs will continue: it extends the addresses sir and Mr to both* genders, except that Cpt Janeway has accrued enough power to carve out an exception for herself.

    * In Star Trek there are strictly two social/biological/electrospiritual genders in the entire galaxy.

  38. https://i.dailymail.co.uk/i/pix/2013/11/14/article-2506952-1963D45400000578-537_634x566.jpg

    Mongolian ambassador wearing blue silk dress as he presents his credentials to the Queen.

    Yes, they are that confident in their masculinity…

  39. That’s not a dress, it’s a deel.

    Still, it’s kind of ironic that the Mongols wear a robe when the Germanic peoples supposedly started wearing trousers because they were more convenient for riding a horse.

  40. David Eddyshaw says:

    It’s one of my favourite examples of cultural conservatism: in the West we wear trousers because the barbarians who overran the western provinces of the the Roman Empire did.

    The Romans thought trousers were effeminate. Or at least foreign. which is much the same thing. Like chairs with backs to them: for the weak.

    My second-favourite example is that we still haven’t come up with anything better to eat off than baked clay.

  41. My eyes tell me it’s a dress. Silk dress.

    Just like kilt is a skirt.

  42. David Eddyshaw says:

    Pshaw! Wittgenstein tells us that it doesn’t matter what it’s made of or what it looks like, it’s all about how it’s used.

    (In this case, to preserve the dignity of a studly ambassador.)

  43. David Marjanović says:

    Mongolian ambassador wearing blue silk dress as he presents his credentials to the Queen.

    I did “specify” Western culture.

    in the West we wear trousers because the barbarians who overran the western provinces of the the Roman Empire did

    But the French still drink their breakfast drinks out of bowls, not out of barbarian mugs.

  44. @David Eddyshaw: That the world is populated by functions was not Wittgenstein’s idea. He got it directly from Husserl.

  45. Stu Clayton says:

    Where did David Eddyshaw say anything about functions populating the world ? Anyway, I surmise you mean functions as operations on phenomena, constituting and reconstituting them. The notion of “world” sticks out like a sore thumb in this context.

  46. @Stu Clayton: Unless I misunderstood what David Eddyshaw is referring to, the idea that, “The world is populated by functions,” is a fairly standard way of expressing the idea he was alluding to. Wittgenstein was, naturally enough, particularly concerned with what that entailed linguistically, but the basic idea (that the most important classification of what something in our environment is is its “authentic function”) is older. I am not sure if Wittgenstein ever phrased it the way I did, but Heidegger (or rather, his translators) certainly did. I used to know all the original German terminology used by Husserl and Heidegger, but, like most of my German vocabulary, my knowledge of the words repurposed and coined to describe concepts in philosophical phenomenology has atrophied a great deal over the last couple of decades.

  47. Stu Clayton says:

    @Brett: the original German terminology of Husserl and Heidegger is hard to save from atrophy. Both of them were hag-ridden, the hags being etymological punning and general neologismania. You have to keep slogging away at their works to sustain a conviction that it all makes useful sense. You must slog for a lifetime.

    In my case I would speak of a heap of collapsed birthday balloons, which are like your withered tails. When someone in the internet seems to be wrong about H&H, I rush to various learnèd sites to pump up the balloons again in my mind. It’s then all very colorful, but somehow the party spirit is gone.

    There are people like Sloterdijk to lead you through Heidegger, and things seem to make sense – as in a dream, until you wake up.

  48. David Eddyshaw says:

    It seems unclear whether LW was actually really acquainted with Husserl’s work: I suspect that insofar as there are parallels, LW was reinventing the wheel. Probably one of those questions we won’t be able to answer definitively until the US military works the bugs out of its time machine.

    http://wab.uib.no/agora/tools/alws/collection-8-issue-1-article-2.annotate

  49. Stu Clayton says:

    Moritz Schlick is mentioned early in that article. Did you know that he was murdered in 1936 by a Trump supporter, whom the Nazis released on probation two years later ?

  50. Stu Clayton says:

    # However, the point is not whether Wittgenstein has been influenced by others, but rather how do these influences manifest themselves in his work, or concerning the question of origin, to what extent can something be regarded as being the source? #

    If the author of that article is a friend of yours, I probably should not have quoted that.

  51. David Eddyshaw says:

    Moritz Schlick

    Yes, I did know: a martyr for Logical Positivism (at least if you take the assassin’s word for it, which you probably shouldn’t.)

  52. Stu Clayton says:

    Remains only to wonder why he was let out on probation after two years of the ten to which he was sentenced. I presume some connection with his stated motives, in the considerations of his probationers:

    Schlicks antimetaphysische Philosophie habe seine moralische Überzeugung verunsichert, wodurch er seinen lebensweltlichen Rück- und Zusammenhalt verloren habe.

  53. David Marjanović says:

    Ah yeah, that’s Nazi logic: Schlick’s philosophy destroyed that guy’s moral compass, so the completely logical reaction is to murder Schlick, because somehow that’ll restore the moral compass. Or at least avenge it, I guess.

    The spot in the U of Vienna where Schlick was murdered is marked on the floor.

  54. John Cowan says:

    I remember him being on the other side of the argument about whether it’s inherently self-contradictory to attend your own funeral or not, but I forget which one held which position. Then again, I was reading the story in one of Smullyan’s books….

  55. David Eddyshaw says:

    According to Wikipedia, after the war the assassin (whose name does not deserve to be remembered) had the crust to sue Viktor Kraft for describing him as a paranoid psychopath. Truly a Trumpian avant la lettre.

  56. John Cowan says:

    I meant to say “on the other side of the argument from Carnap”. A quick check shows that Schlick took the affirmative, Carnap the negative.

    does not deserve to be remembered

    Ah yes, the old Herostratus Trick. But that never works!

  57. John Cowan says:

    If Semantic Primes are supposed to exist independent of any particular actual language

    It depends on what you mean by “independent”. They are certainly not a priori; the list grew from 14 to 65 when more languages were investigated. If a word in some natural language is not expressible using the primes, then it is a new prime; they are (metaphorically) mathematical prime numbers rather than primitive notions.

    A conceptual difficulty is that the primes may be polysemous in any given natural language, but only one of those meanings is intended: SEE is what you do with your eyes. But of course there is no precise specification of either how many meanings a mot has[1] or which one of them actually is the prime. We just have to fall back on the fact that TWO has a central meaning that everyone who speaks not only English but almost all languages can agree on (Pirahã perhaps being an exception), unlike child which is culture-bound and requires an explication, as do Kind and enfant and so on. But if we could explicate TWO in universal terms, it wouldn’t be a prime.

    The framework of NSM, like generativism or lowercased optimality theory, is not empirical and can’t be called true or false, only more useful or less useful. (I’ve always admired the way in which Le Guin calls certain translations of the Daodejing “not useful”.) So particular NSM theories (i.e. some particular set of primes) has been refuted frequently, but the framework can only be neglected or abandoned.

    I don’t know if Ewe is one of your languages, but you can download the semantic primes in Ewe (Microsoft Word format, but can be read by OpenOffice and LibreOffice) and many other languages (but regrettably not Hausa).

    [1] As Quine put it, there is no place in science for ideas, no place in epistemology for knowledge, and no place in semantics for meanings.

    All the semantic baggage that [articles] bring is strictly secondary.

    Secondary to Russian-speakers, that is, but essential to English-speakers. In Mandarin, the difference between present tense and past tense is secondary (it need not be expressed when it’s obvious), but the imperfective/perfective distinction is mandatory, as in Russian, but in English this p/i distinction is secondary and seems quite unnecessary most of the time.

    Let’s take a simple Russian sentence “Картина висит на стене как-то косо”. Word for word English is “Picture is hanging on wall somewhat askew”.

    In English, the differences between “The picture is hanginag on the wall” (I expect that you know which ones I mean), “A picture is hanging on the wall” (I think you know which wall but not which picture), “The picture is hanging on a wall” (a little odd pragmatically, because it’s not usual for me to assume you know which picture I’m talking about but not know which wall) and “A picture is hanging on a wall” (of course it is, and so for many pictures and many walls) are fundamental, and English-speakers cannot imagine not expressing them.

  58. The picture is hanging on a wall

    Yes, slightly odd pragmatically, but could be expanded to:

    The stolen picture is hanging on a wall in some Nazi leader’s mansion.

    Or:

    The picture hasn’t been relegated to storage somewhere; it’s hanging on a wall

  59. Sorry, but I am not convinced. For example, it’s very difficult in Russian to talk about someone in the 3rd person singular and not give away their gender. Russian simply is not designed to maintain this uncertainty. But it is relatively easy to pull off in absolutely naturally sounding English. Because I unconsciously expect to know the gender of a person being described, I found myself many times ascribing (without substantial evidence) a gender to people I read about and coming out the wrong way. In this case, Russian grammar forces certain information being revealed even if there is no actual need or desire to do so. English articles are similar, you cannot simply refuse to use them if you really don’t care whether an unknown picture is hanging on the wall or a well-known picture is hanging on the wall. Of course, in most situations Russian speakers know perfectly well what sort of picture is that. From the context.
    That makes me think that English articles play a grammatical role first and semantic role second.

  60. John Cowan says:

    But there are also situations where the article is omitted, and that means something different. It’s not like French where you pretty much can’t have a noun phrase without an article. As I’ve posted before, Elvis was in a jail (physically inside one) when he made “Jailhouse Rock”, but he was never in jail (imprisoned).

  61. For example, it’s very difficult in Russian to talk about someone in the 3rd person singular and not give away their gender. Russian simply is not designed to maintain this uncertainty.

    Strugacky brothers managed to do that in “The Dead Mountaineer’s Inn“.

  62. Stu Clayton says:

    What is this discussion about, now we’re at half-time ? Ways to improve one language on the model of another ? One subtheme seems to be frustration at an inability to feel at home in certain foreign languages, because reasons. Another seems to be the heady delights of flogging dead horses, such as Semantic Primes.

    At times like these, I find consolation in meditating on the fact that Sinn ist unentrinnbar and addictive. It’s hard to ignore. Otherwise I wouldn’t be griping.

  63. Russian grammar forces certain information …

    See and hear Plungyan speak about it (in Russian):

    https://postnauka.ru/video/1898

  64. Stu Clayton says:

    The very notion of “forcing certain information” is a child of current times. It is suffering from genderitis, resentfully unsung sexualities and data-protection laws. Until about 30 years ago, few people made a stink about “he” and “she”. Now the fuss couldn’t be wuss.

  65. David Eddyshaw says:

    The framework of NSM, like generativism or lowercased optimality theory, is not empirical and can’t be called true or false, only more useful or less useful.

    Thanks for the clarification, JC: that was the impression I had got. I suppose my objections could be said to boil down to the ancient French intellectual complaint: “Very well: it may work in practice. But does it work in theory?”

  66. Stu Clayton says:

    But that is the complaint of scientists schlechthin, is it not ? Unless one knows the mechanisms of an effective practice – how it works, the “theory of it” – one is not entitled to claim that it is effective, or even to allow it to be effective. Last year I learned that this was a main objection to Semmelweiß. He discovered a correlation between unsanitary doctors’ hands and outbreaks of puerperal fever. He had no theoretical explanation for it, or only one that seemed crazy at the time, and so was laughed to scorn.

    A similar debacle was the reaction of the sci-comm to that Australian’s (?) discovery of heliobacter pylori.

    This is a downside of an attitude that otherwise praiseworthily keeps peach-pit extract treatments of cancer at bay.

    Are we to treat the fuckups as “statistically insignificant” ? Is there a sound theory of statistical Right and Wrong, applicable to science in general ? Or is one guided here by pragmatics tricked out as mathematics ?

  67. What is this discussion about, now we’re at half-time ?

    To me it seems to be trampling over reasonably well-tilled soil and returning it to primeval swamp. Just stop. @Bathrobe that wiki on ‘Specificity’ is a stub and dreadful and a waste of time. Is this the best Linguistics can do? I suggest the Stamford Encyclopedia of Philosophy’s article on (definite) ‘Descriptions’.

    I’m happy to call any tool for semantics (like NSM) “less useful” unless it can account for anaphora — as is one of the functions of ‘the’ (or ‘this’) in English: there’s a quantifier somewhere in the context of utterance; there’s references back or forward to whatever’s quantified. Any account of anaphora must be able to explain why this is funny: “You saw noone, and at that distance? You must have good eyesight!”

    Logicians do not claim that what is semantically prime about quantification is the quantifier: it’s the whole frame over which the quantified variable is scoped, including the syntactic appearance of the quantifier. I don’t see that any theory of semantics which attaches meaning purely to appearances of words is going to be adequate.

    Re JC’s claim about that Russian sentence: whilst grammatical, I plain don’t believe that is unproblematic for Russian speakers if somebody burst into a theatre, declaimed it, then rushed out without another word. Would hearers assess it as true or false? Would they all agree at least how to go about assessing its truthhood? That is, without more context of the utterance. (Oh, and presuming the declaimer didn’t point at or glance at some picture or wall.) Truthhood is not everything in semantics, but if you can’t even agree on that, you’re not at first base.

  68. David Eddyshaw says:

    I like Plungian. I fist came across him in the context of work on African languages: specifically a highly illuminating paper on what he calls “discontinuous past” constructions, which are sort of anti-current-relevance pasts with the specific implication that the events described are done-and-dusted. Kusaal has a today-past discontinuous past, as I had laboriously teased out on my own; it was very helpful to discover from P’s work that this was no isolated aberration, especially in Africa.

    Also, Vladimir Plungian is a cool name. It just is.

  69. Plungyan (פלונגיאן) is a Yiddish name of the Lithuanian town of Plongė (formerly Plungyany in Russian).

    The town used to have predominantly Jewish population until July 15, 1941.

  70. David Eddyshaw says:

    @SF: Thanks. I’d wondered about the origin of the name.

  71. David Eddyshaw says:

    Looking at the Wikipedia article on him, I see that his father was a friend of Yuri Knorozov, as in Maya.

    Incidentally, the ‘pedia picture of Knorozov with his cat is wonderful.

  72. David Eddyshaw says:

    @Juha:

    For the benefit of us poor souls without Russian, does Plungian’s talk on lexicon lend itself to summary?

  73. David Marjanović says:

    Are we to treat the fuckups as “statistically insignificant” ? Is there a sound theory of statistical Right and Wrong, applicable to science in general ? Or is one guided here by pragmatics tricked out as mathematics ?

    Well, you arbitrarily pick a p value which you decide to treat as “significant” (usually 5% for sheer lack of imagination); then, when the p that the fuckups are random drops below the value you picked, you declare the fuckups statistically significant. If you’re good, you declare them statistically significant at the p = 0.05 level (…what a compound noun).

    BTW, Helicobacter is the one in stomach ulcers; heliobacteria are photosynthetic. And the good man is spelled Semmelweis just to confuse people.

  74. PlasticPaddy says:

    @dm
    Re Semmelweis the weis bit is because the rolls were wheaten, not necessarily made from white flour. Compare Weißbier. I do not understand why the sz was not used in Hungarian spelling of this name, did they pronounce it Semmelweisch there?

  75. Stu Clayton says:

    when the p that the fuckups are random drops below the value you picked

    But the fuckups and their adversaries are states of affairs which can’t be summoned up by the magic of random trials. That’s an example of what I meant by “pragmatics tricked out as mathematics”. I’m just as pragmatic as the next guy, my question is merely (as David Eddyshaw put it for unnamed French intellectuals) “does it work in theory ?” You can tally and compare occurrences of each kind, but that’s about it as far as the math goes.

    Semmelweis. Helicobacter.

  76. Plungian on lexicon:

    Languages have words, but are they the same words in all languages? It’s in lexicon that languages differ most dramatically. This has long been noticed and some conclusions drawn. First of all, it would be nice to know what kind of difference it is. For example, some languages have two words for the upper limb, eg ‘hand’ and ‘arm’ in English, but others have only one, eg ‘рука /ruka’ in Russian. To be sure, Russian does have ‘кисть /kist’ to refer to the hand, but lacks a word to refer to the upper limb from the shoulder to the wrist. A hypothesis has even been put forward that two words are characteristic of parts of the world where people wear clothes so that only the hands are visible. In warmer climates, the entire upper limbs are visible, so there is no need for two words. However, the Slavic languages disprove it: they have only one word and the Slavic people have always lived in parts of the world with a temperate climate. Likewise with parts of the day. All languages have words for sunrise and sunset, but everything else may differ greatly.
    Colour words are a favourite example. Russian has seven basic colour words, but not all languages have exactly that number. Some have just two or three words, others four or five, and still others have many more than seven. Some languages may distinguish light and dark colours, and the variety is enormous. Kinship terms are another area of great variety. In addition to ‘брат /brat (brother)’ and ‘сестра /sestra (sister)’, there may be terms for ‘elder brother/sister’, ‘younger brother /sister’. Some languages do not distinguish sex/gender, and employ words to the effect of ‘same-sex elder/younger child of the same parents’ and ‘opposite-sex elder/younger child of the same parents’. Every language is a very crude snapshot of the world. No language can reflect everything in its infinite variety, and the lexicon of a language has to make choices.
    When it come to abstract lexicon, the differences become even greater, such as terms to refer to emotions, moral qualities, etc. Take words for ‘love’. To love is human, but the big question is how to refer to it. Some languages have no special word for ‘love’, and the verb for loving is the same as for wanting, as ‘querer’ in Spanish. There are even languages with no word for ‘wanting’, but they are a minority. An example to the contrary is ancient Greek, which has at least four words: έρος, φιλία, αγάπη, and στοργή. Those are everyday ancient Greek words. They had an influence on ancient philosophy and Christian ethics. And such different ways of carving up the world are found in everywhere in the lexicon.
    Nouns and verbs are not universal categories either. Some languages may not have adjectives, prepositions, and employ some other parts of speech. Languages may have highly exotic ways of describing the world. A language in the north of Russia has a verb that means ‘lying down and watching others do work’. By the way, Russian has a similar verb, ‘prokhazhdat’sya’ (=loaf about).
    Faced with such diversity, the question arises: what to do with such various ways of taking snapshots of the world. Philologists have often run into extremes. Since every language has a unique worldview, people speaking different languages must think in different ways (the Sapir–Whorf hypothesis). The differences between languages are staggering, they affect the way we think, so we cannot understand each other. At the level of thinking, people are much closer to each other than at the level of language. But the gap between the two is not as unbridgeable as some tend to think, but neither is it as insignificant as some universalists /globalist would have it. People are different, and we should preserve and study such differences: we will only be better for it.

  77. The very notion of “forcing certain information” is a child of current times. It is suffering from genderitis, resentfully unsung sexualities and data-protection laws. Until about 30 years ago, few people made a stink about “he” and “she”. Now the fuss couldn’t be wuss.

    You couldn’t be wronger, and once again you are applying entirely irrelevant hobbyhorses to simple linguistic facts. “Forcing information” has nothing whatever to do with feminism or gender identity, it is a matter of what a language forces you to express, which can be gender, number, status, or a million other things, and I assure you it has been a matter of interest since the very beginning of linguistics.

  78. Ok, maybe “forcing” is too strong (never thought that Mr. Clayton is so sensitive). Most people in most situations do provide this information willingly, happily, and without noticing. The trouble begins when they specifically do not want to reveal it either out of caginess or because they don’t know. A really good example pointed out by SFReader ( “The Dead Mountaineer’s Inn“) shows precisely how awkward the situation is when that happens.

  79. David Eddyshaw says:

    From Sapir’s Language, discussing variants and versions of the sentence “The farmer killed the duckling”:

    The Yana sentence has already illustrated the point that certain of our supposedly essential concepts may be ignored; both the Yana and the German sentence illustrate the further point that certain concepts may need expression for which an English-speaking person, or rather the English-speaking habit, finds no need whatever. One could go on and give endless examples of such deviations from English form, but we shall have to content ourselves with a few more indications. In the Chinese sentence “Man kill duck,” which may be looked upon as the practical equivalent of “The man kills the duck,” there is by no means present for the Chinese consciousness that childish, halting, empty feeling which we experience in the literal English translation. The three concrete concepts – two objects and an action – are each directly expressed by a monosyllabic word which is at the same time a radical element; the two relational concepts – “subject” and “object” – are expressed solely by the position of the concrete words before and after the word of action. And that is all. Definiteness or indefiniteness of reference, number, personality as an inherent aspect of the verb, tense, not to speak of gender – all these are given no expression in the Chinese sentence, which, for all that, is a perfectly adequate communication – provided, of course, there is that context, that background of mutual understanding that is essential to the complete intelligibility of all speech. Nor does this qualification impair our argument, for in the English sentence too we leave unexpressed a large number of ideas which are either taken for granted or which have been developed or are about to be developed in the course of the conversation. Nothing has been said, for example, in the English, German, Yana, or Chinese sentence as to the place relations of the farmer, the duck, the speaker, and the listener. Are the farmer and the duck both visible or is one or the other invisible from the point of view of the speaker, and are both placed within the horizon of the speaker, the listener, or of some indefinite point of reference “off yonder”? In other words, to paraphrase awkwardly certain latent “demonstrative” ideas, does this farmer (invisible to us but standing behind a door not far away from me, you being seated yonder well out of reach) kill that duckling (which belongs to you)? or does that farmer (who lives in your neighborhood and whom we see over there) kill that duckling (that belongs to him)? This type of demonstrative elaboration is foreign to our way of thinking, but it would seem very natural, indeed unavoidable, to a Kwakiutl Indian.

  80. John Cowan says:

    I fist came across him in the context of work on African languages

    I see how you nounincorporated that.

    He had no theoretical explanation for it, or only one that seemed crazy at the time, and so was laughed to scorn.

    The same can be said of Alfred Wegener (I checked the spelling). “What do you mean the continents are crossing the oceans! How? And why doesn’t the oceanic rock pile up on one side?” And his explanations were hopeless: “Moved by magnetism? Really???” Only when a plausible mechanism turned up through quite independent lines of research did earth science divert itself, like the Mississippi roaring down the Atchafalaya and leaving New Orleans and Baton Rouge dry. (This hasn’t in fact happened, thanks to the dam complex at the confluence with the Red River that limits how much water the Atchafalaya gets.)

    Languages have words, but are they the same words in all languages?

    NSM not only replies “No!” but gives a mechanism for explaining what all those words actually do mean in a way that can in principle be translated into any language. Pace AntC, it is not meant to be a theory of semantics as a whole, only a theory of word meanings. (As far as I am concerned, any theory of semantics that excludes pragmatics ends up being a theory about what discourse does not mean.)

    never thought that Mr. Clayton is so sensitive

    Not he! He just plays one on TV; the self-assigned nickname “Grumbly” is appropriate.

  81. Stu Clayton says:

    D.O., I am not motivated by emotional sensitivity, if that’s what you mean. I am, rather, greatly irritated by this false idol of the study called “forcing information” or perhaps “requiring information”.

    Languages neither reveal nor conceal information, since they are not motivated to reveal or conceal. People have these motives.

    You say “most people in most situations do provide this information willingly, happily, and without noticing”. The crucial bit of that is “not noticing”. If a speaker of language X has some information he wants to conceal, he finds that X provides plenty of techniques to do that – and if not, in a given case, he can invent one. One obvious technique is the “null way” – just not say anything on the subject, and it will be concealed automatically, by not being revealed. Of course things are not quite that easy – persons (present or absent) paying attention to what the speaker says or writes may conclude, from the speaker’s failure to speak up in connection with a topic they expect him to speak up on, that he is concealing something about that topic.

    The speaker of X also can supply any information he wants to supply – provided only that those listening understand as information what X says. The language X provides plenty of techniques to do that – and if not, in a given case, the speaker can invent one, say with a neologism.

    Now if the speaker of X is learning a language Y, or perhaps has already learned it, he may have occasion to compare X and Y like this: “Using Y I have to reveal information that I don’t have to reveal using X”. One example cited above was gender, that’s why I mocked it. I don’t give a damn about unsung sexualities.

    Why doesn’t it occur to the speaker to analyze the concealing/revealing resources of X for what they are, and those of Y for what they are, and just leave it at that ? Could it be that he doesn’t speak Y as well as he would like, so faults it for not being like X ?

    I speak, read, write and understand English and German with equal proficiency. It took a long time with German, but here I am to challenge all comers, just as in English. I have NEVER, in either language, experienced any “inherent” difficulty concealing or revealing information – because I know the rhetorical ropes in each. Periphrasis, insinuation, lying – all useful techniques that are missing in neither language. I ALWAYS find the explanation for occasional difficulties in myself, or in the lack of ready-made traditions of writing about certain things in certain ways (see the difficulties in conveying Luhmann’s prose in English).

    Only when confronting what’s called “translation” between English and German, can one get the idea the English forces etc and German doesn’t, or vice versa etc. The most frequent reason for that, I suspect, is being hobbled by notions of needing to match noun counts, verb counts, be faithful to the spirit and the letter and so on.

    The idea of “forcing information” is an illusion induced by “translation”. I say it’s spinach, and I say the hell with it.

    Edit: David Eddyshaw above gave a very good commentary on these matters, in a more moderate tone. He will be remembered as a mover of men, while I will go griping to an ignominious grave.

  82. The idea of “forcing information” is an illusion induced by “translation”. I say it’s spinach, and I say the hell with it.

    You have no idea what you’re talking about, but don’t let that stop you. Languages require certain information systematically and without conscious input from speakers.

  83. David Eddyshaw says:

    @Stu:

    I dunno. While it’s always possible to be vague, it can be easier in some languages than others, depending on just what you’re trying to be vague about.

    In the Kusaal Bible, when the translators came to references to Jesus’ brothers, they could happily use a word which indifferently means “sibling” or “cousin” and ignore any doctrinal difficulties; on the other hand, you have to pick whether you mean older or younger sibling/cousin in Kusaal. There are compound words which cover both, but then you have to decide between “same-father” or “same-mother”: you could explicitly specify both, but then the Inquisition would come and get you.

    In fact, they just went with pitib “younger siblings/cousins of the same sex”: and who shall blame them?

  84. Stu Clayton says:

    An observer may see “information being required” from a speaker. A speaker does not feel that “information is being required” from him. As D.O. put it, the speaker doesn’t notice – and if he does and it thwarts his purposes, he can circumvent that requirement.

    An observer can know things about the speaker’s behavior that a speaker might not know, but might be taught to know – but perhaps not to care. A speaker may know things about his behavior that an observer cannot know, but might be taught to know – but perhaps not to care, if the observer has not noticed that knowing works both ways.

  85. Stu Clayton says:

    @David: are those not most simply understood as cultural differences in “what counts”, differences reflected, unsurprisingly, in the languages ? You don’t have to go the whole Sapir-Whorf mile. I imagine translating Grimms’ fairy tales, or Peyton Place, into Kusaal is no easier than translating the bible, even without Inquisitors glaring in the background.

    This whole discussion, and others like it, is vitiated by the absence of an adequate theory of communication and sense. Of course anyone can guess my opinion as to where that, or rather an honest attempt at it, is to be found.

  86. David Eddyshaw says:

    are those not most simply understood as cultural differences in “what counts”, differences reflected, unsurprisingly, in the languages

    Well, yes: that’s pretty much what I actually do mean (and why I was earlier lamenting the unmooring of linguistics from anthropology perpetrated by Some People in the last century.) I don’t go for the Sapir-Whorf thing at all (but then, neither did Sapir or Whorf, as far as I can make out.)

    The Bible example just lay to hand; I didn’t bring it up for any particular ideological reason. Peyton Place is probably untranslatable into any human language, but the Grimms’ stories are often pretty easy to transpose into African modes: unsurprising, as a good few of them are just European representatives of stories told right across the Old World. One of my favourite exhibits of that sort of thing is a Kusaasi folk story with the same plot as Chaucer’s Pardoner’s Tale; both go back in fact to the same Buddhist Jātaka tale.

    The Kusaal “sibling” words can’t by any stretch be interpreted as having imposed a particular world-view of sibling relationships on the Kusaasi: one of the striking oddities of kinship words in Western Oti-Volta is that not only are the various terms for siblings not cognate between very closely related languages like Kusaal and Mampruli, but the systems don’t even have the same categories. Weird.

  87. Stu Clayton says:

    That reminds me of a cute circumstance perhaps exploited by Luhmann in discussing two main ways of responding to disappointed expectations (of any kind) – cognitive and normative. “Cognitive” means “willing to change my attitude”, “normative” means “I’ll stick with my attitude”. The latter occurs, for example, in adherence to moral principles: “would you not kill to protect your sister being sexually attacked by a German” ? “No, I would try to interpose myself”, so the term is not inherently deprecatory.

    However, he also explicates the two terms as lernfähig and lernresistent. In everyday speech nowadays, you can describe someone sarcastically as lernresistent, meaning he’s dumb or as stubborn as a mule. I would dearly love to know whether this colloquial usage preceeded Luhmann, or derives from him, or is all a terrible misunderstanding. Much as an und für sich seems to be an expression that turned up in 18C German philosophy (in se, per se …), and in particular was exploited to the hilt by Hegel. Now it’s just an everyday expression meaning “essentially”.

    Edit: you may well ask in what way what you said reminded me of this. I knew when I started writing it, but now I can’t reconstruct the connection. Maybe “cognate” and “weird” suggested cognition, and who will blame them ?

  88. David Eddyshaw says:

    I would like to think of myself as lernresistent: after all, with attitudes as refined and elevated as mine, it would be positively criminal to alter them in response to any mere fact. What kind of example would I then be setting to those less fortunately endowed?

  89. Stu Clayton says:

    Just so, and I feel the same. One of the fun things was acquiring all the propitiatory techniques that facilitate social interaction, advisable even in one in my position.

  90. I don’t like the English word normative in that meaning either, although I find it harder to express why. I once had a conversation with some philosophy of science researchers about whether results from the path integral formulation of quantum mechanics were ever normative.

  91. Stu Clayton says:

    In turn, what could “normative” mean as applied to the results of a particular formulation of quantum mechanics ? Does it mean exemplary, or obligatory ? Where are the enforcement officers ?

  92. Does anyone know of some theoretical approaches to semantics other than Natural Semantic Metalanguage? What are some of the major differences between Natural Semantic Metalanguage and other approaches?

  93. @Stu Clayton: The issue was what happens when there are potential disagreements between the apparent results of a path integral formulation and a canonical quantization formulation of a model theory. The path integral is, mathematically, very hazily defined, and I argued that results from the path integral would never be considered normative if there seemed to be a difference between the two approaches. What that meant was that in a situation like that, the normal course of action would be to take the canonical results as guidance and try to figure out how to set up the path integral to get the same result.

  94. @AntC

    Is this the best Linguistics can do?

    No, it’s Wikipedia. What do you expect?

    I suggest the Stamford Encyclopedia of Philosophy’s article on (definite) ‘Descriptions’.

    Thank you for finding such a detailed discussion. I will read it thoroughly at my leisure. There is an awful lot of linguistics in there, much of it concerned with the facts of English usage. Specificity gets only a brief (and from the point of view of my examples) irrelevant look-in.

    While this kind of problem appears to be of great interest to logicians, it is of equal interest to linguists. Towards the end there is a discussion of determiners as case markers (based, as usual, on a lamentable paucity of languages), but the question of “possession” is interesting. There are languages (like Mongolian), where the notion of possession has been extended to become a major anaphoric principle, beyond the ordinary sense of possession.

    The uses of the definite article in English are, in fact, quite complex. As far as I know, no linguist, let alone logician, has managed to explain them completely. If there were a case where linguistic usage disagreed with the logician’s edifice, which would take precedence?

    Any account of anaphora must be able to explain why this is funny: “You saw noone, and at that distance? You must have good eyesight!”

    Actually, also an account of negation.

    Logicians do not claim that what is semantically prime about quantification is the quantifier: it’s the whole frame over which the quantified variable is scoped, including the syntactic appearance of the quantifier.

    Substitute ‘linguists’ and the statement is equally true.

    if somebody burst into a theatre, declaimed it, then rushed out without another word

    This is an extreme example, the kind that is frequently concocted in an attempt to pare away any semblance of ordinary context and get at the ‘heart’ of the phenomenon. The results are frequently absurd.

    To be honest, I don’t quite understand why you are so angry.

    @LH and Stu

    LH is right about the “forcing” part, but Stu’s example is apt. Whatever grounds people may have for it, the push against “he” and “she” does represent an attempt to change the kind of information that is required to be given in the course of “speaking English”.

    @Stu

    would you not kill to protect your sister being sexually attacked by a German

    Would this extend to killing your sister?

  95. Kinship terms:

    These are a favourite in anthropology and linguistics. They can also, as David Eddyshaw pointed out, be the bane of the translator.

    There is an interesting example in Harry Potter. As fans know, Harry was adopted into his aunt’s family, who subjected him to various forms of mistreatment, which is most evident in the favouritism shown towards his ‘step-brother’ (that is, cousin) Dudley.

    One minor example of this is that Harry was forced to call his uncle’s sister ‘Aunt Marge’.
    Dudley had two aunts: Aunt Lily (aunt on his mother’s side, killed by Voldemort), and Aunt Marge (on his father’s side). ‘Aunt Lily’, never mentioned in the family, was actually Harry’s mother. ‘Aunt Marge’, on Dudley’s father’s side, has no kinship relationship to Harry.

    This is something of a trap for careless translators. A couple actually mistranslate ‘aunt’ in ‘Aunt Marge’ in such a way as to refer to Dudley’s mother’s younger sister — Harry’s mother — rather than Dudley’s father’s elder sister.

    (See my web page Aunt Marge’s Big Mistake.)

  96. David Marjanović says:

    I do not understand why the sz was not used in Hungarian spelling of this name

    Evidently the spelling is kept completely German. I don’t think the homepage of the Semmelweis Museum says anything about pronunciation.

    If there’s a reason for the -s, as opposed to the usual elegant variation, I can offer phonetics: I pronounce weiß with a short /s/, but weiße with a long one.

    Now it’s just an everyday expression meaning “essentially”.

    Worse – the /d/ has dropped out, and so it took years for me to notice that the whole thing was parsable and wasn’t just a single word with penultimate stress.

    And only just now have I figured out that the stress is contrastive.

  97. Plungyan: Russian has seven basic colour words

    How does he reckon? Seven colors of the rainbow? Red-orange-yellow-green-light blue-blue-violet? Ignoring obvious words for black, white, and gray, pink and brown beat orange and violet in frequency. I don’t see how it can be seven.

  98. LH is right about the “forcing” part, but Stu’s example is apt. Whatever grounds people may have for it, the push against “he” and “she” does represent an attempt to change the kind of information that is required to be given in the course of “speaking English”.

    Stu’s example is not apt, it shows he has no idea what is being discussed. He leaps automatically to what is salient in his mind, the presumed excesses of gender correctness, without realizing that none of that is relevant to the topic of what grammar forces you to say except that the proposed change in pronouns is intended to alter a small corner of the vast carpet of grammatical information. It’s like the “Fire!” example.

  99. Stu Clayton says:

    Is a snake forced to go on its belly, since it has no legs ? No, that is merely the way a snake proceeds. Its autonomy is not impaired.

    Is a chess player forced to prevent his king from bring captured ? No, he is merely following the rules when he does so.

    Is a heliotropic plant forced to turn towards the sun ? No, that’s just what it does.

    Is an English speaker forced to provide information he would rather not reveal, such as gender ? No, he’s just following the rules of speech about persons when he provides such information – and he can avoid that to a certain extent.

    The notion of “forcing” is teleological, unless you admit that it is being used metaphorically here with respect to regularities of speech. And if you admit that, you must admit that it is only a metaphor, and does not imply force in any useful, forceful sense.

  100. PlasticPaddy says:

    @stu
    I am not sure force requires a forceful sense. Maybe constraint would be a more descriptive term, but force is the one linguists use (the term “centrifugal force” is used commonly for something which is not a “true” force, but the result of a constraint). If I understand correctly, what they are talking about is something like the fact that you go up to Dublin from anywhere in Ireland but if you want to use a word corresponding to “up” in Irish, you are “forced” to provide the main cardinal direction of travel (this is true even for line of sight trips).

  101. @Stu: Qui potest mori non potest cogi?
    But I have seen “is forced” being used in those cases where you say it doesn’t apply, or cases very similar to them, so that is part of the usual scope of meaning of the word.
    (EDIT: Ninja’d in a way by Paddy.)

  102. Stu Clayton says:

    @Hans, @Paddy: I understand that linguists use the word “force” as they do. My point is that the use of this word is worse than unfortunate, since it foments a concern with issues (“empowerment” etc) not relevant to speech regularities. It’s also unnecessary, since you can formulate everything that needs to be said in terms of regularity, rules and rule-following behavior.

    It’s a small point, one of many that can and have been made about the deleterious effects that incautious terminology can have on thought.

    Consider this: “the purpose of wings in birds is to enable flight”. In a way, a harmless way of speaking if you are willing to put it less teleologically. Creationists are not willing. Only slightly better would be “the function of wings in birds is to enable flight”, since function does not imply purpose – provided one has learned to distinguish between them in that way.

    Cogi possum, ergo chump.

  103. Stu Clayton says:

    If languages “forced” people to speak in certain ways, all the unsung sexualites and empowerment people would have to close shop. But there is no force, rules and practices can be changed and they are being changed. That is not a bad thing per se. Unless they’re forced on you …

  104. David Eddyshaw says:

    Aunt Marge

    In Kusaal, Amarisi would just have to be Ayaba’s (Aya’ab “Potter”) makpɛɛm sid pu’a: makpɛɛm “mother’s elder sister or mother’s senior co-wife”, a compound of ma “mother” and kpɛɛm “elder”, here construed adjectivally; sid “husband”; pu’a “woman, wife.”

    One of my informants, who spoke almost perfect English, found it so conceptually bizarre that English uses “aunt” for people who are not blood relations that he asked me whether the term had something to do with seniority within the family (the guiding principle of the Kusaasi system.) Couldn’t get his head around it.

  105. Stu Clayton says:

    Don’t the Kusaasi (?) have honorific family terms for each other, such as “father” in Turkish ? I hear it so much here in Cologne that every child must be father to every other. Is it that women are not accorded such terms ?

  106. I think I mentioned that here before, but anyway. When I was a child, about 40 years ago, it was still usual for children in Germany to address adults, except for their parents / grandparents, with Onkel / Tante, independent of whether they were related or not. With some friends of my parents I only learnt as a teenager that they weren’t actually relatives. (Also, I don’t have any uncles or aunts sensu strictu, because both my parents are only children.)

  107. PlasticPaddy says:

    @de
    I think the Kusaasi are on to something. How can we otherwise explain the power Bertie Wooster”s aunts have over him (and the terror they inspire in him)? Although Bertie is to some extent financially beholden to these aunts, and although they may indeed be blood relations (on this point I am not clear) these considerations would not seem sufficient in themselves.

  108. I wonder what the Kusaasi would make of the Spanish colloquial “tía”

  109. David Eddyshaw says:

    You might politely call an unrelated woman old enough to actually be your mother, ma, and a much older man, ba’; oddly enough, you couldn’t use the formal word for father, saam (unless he actually was your father.)

    Kusaasi don’t use their “uncle” and “aunt” words as respectful terms to strangers; instead, they use diem, which as a family relationship term refers to a man’s parent-in-law of either sex. As a respect term, it is used by both sexes, and applies to any unrelated person of opposite sex and about the same age as oneself or older but not old enough to be called ma or ba’. For a person of similar age and same sex you’d use zua “friend”, and to anyone much younger, biig “child.”

    It’s perfectly polite to address someone by a generic term, like: Pu’asadirɛ! “Young-woman-who-hasn’t-yet-given-birth!”; Nasaara! “White man!”

    The Kusaasi don’t really do hierarchy-type respect words. They didn’t have chiefs until they were imposed on them by outsiders, and have never really internalised the deference bit.

    In the Kusaal Bible, In the resurrection story in John’s Gospel, when Mary Magdalene initially fails to recognise Jesus, she addresses him as m diemaa “my in-law.”

    Thinking about it, I suppose this use of diem is not so much a respect term as an avoidance term: in general, there are no particular rules about polite address of strangers, but when it comes to talking to someone you might in principle have a sexual relationship with, you need to have a more distancing term available (parents-in-law are greatly respected in the local culture, though you have a mutual joking relationship with your sibs-in-law.) That’s just a vague anthropological speculation on my part; the speakers simply regard the diem of respect as the “same word” as the relationship term, as I discovered from their attempts to explain the usage to me.

  110. Lars Mathiesen says:

    Danish, in many families at least, separates grandparents by relation, aunts and uncles additionally by whether they married into the family, and cousins (always) by gender:

    Mother’s mother: mormor (optionally bedstemor)
    Father’s mother: farmor (optionally bedstemor)
    Mother’s father: morfar (optionally bedstefar)
    Father’s father: farfar (optionally bedstefar)

    Mother’s sister: moster (optionally tante)
    Mother’s brother: morbror (optionally onkel)
    Father’s sister: faster (optionally tante)
    Father’s brother: farbror (optionally onkel)
    Wife of parent’s sibling: tante (always)
    Husband of parent’s sibling: onkel (always)

    Female cousin: kusine
    Male cousin: fætter

    Swedish doesn’t have the equivalent of bedstemor/far, and kusin covers both genders. (Some sources indicate that moster and so on can cover the tante/onkel space too, in contradistinction to Danish, but I never discussed family trees in detail with actual Swedes so I’m not sure).

    So even between Danish and Swedish you can get that semantic lacuna — to my Danish mind, male and female cousins are too different to lump together.

  111. The notion of “forcing” is teleological, unless you admit that it is being used metaphorically here with respect to regularities of speech. And if you admit that, you must admit that it is only a metaphor, and does not imply force in any useful, forceful sense.

    Yes, all of language is “only a metaphor” if you delve deep enough. If it amuses you to point this out in random individual cases, by all means knock yourself out, but it is absurd to say “the use of this word is worse than unfortunate, since it foments a concern with issues (’empowerment’ etc) not relevant to speech regularities.” The only one in whom such a concern is fomented is you, and I am forced to assume you are fomenting your own fomentation for the sake of your amusement, since you cannot possibly seriously think that anything hangs on the word choice of linguists.

  112. In the Chinese sentence “Man kill duck,” which may be looked upon as the practical equivalent of “The man kills the duck”.

    Chinese would use either nánrén shā yā (男人殺鴨) ‘male person kill duck’ or rén shā yā 人殺鴨 ‘person kill duck’.

    Aunt Marge in Kusaal:

    I think Aunt Marge would have to be Ayaba’s (Aya’ab “Potter”) “father’s elder sister or father’s senior co-wife”. But, of course, Aunt Marge isn’t Harry’s aunt, she is Dudley’s.

  113. David Eddyshaw says:

    Ah. My unfamiliarity with the canon betrayed me into error.

    Amarisi is of course Ayaba’s saam dakitua taʋn “father’s wife’s-sister’s-husband’s sibling-of-opposite-sex-regardless-of-seniority.”

    Apologies to all Kusaasi fans of Ayaba. I shall try harder in future.

    person kill duck

    I think Sapir was using “man” for 人; this was 1921, after all. And he does say “not to speak of gender.”

  114. You might politely call an unrelated woman old enough to actually be your mother, ma, and a much older man, ba

    Mma Ramotswe [Precious Ramotswe in Alexander McCall Smith’s No1 Ladies Detective Agency] is a Motswana woman [and] the main detective. Mma is a Setswana term of respect for a woman; the equivalent term for a man is Rra – Wiki.

    both my parents are only children
    They probably said the same about you.

  115. Well caught.

  116. David Eddyshaw says:

    I think Aunt Marge would have to be Ayaba’s (Aya’ab “Potter”) “father’s elder sister or father’s senior co-wife”. But, of course, Aunt Marge isn’t Harry’s aunt, she is Dudley’s.

    Amarisi is Aradiri’s pʋgʋdib “father’s sister” (seniority irrelevant: the system only marks it between relatives of the same sex,)

  117. Athel Cornish-Bowden says:

    I do not understand why the sz was not used in Hungarian spelling of this name, did they pronounce it Semmelweisch there?

    I don’t understand that either. I went to a meeting in Debrecen in 1985 and I could swear that I remember some connection of the Medical School with Semmelweis, but if my searching on the web is to be believed I’m wrong about that, as pages about Semmelweis don’t mention Debrecen in any useful way, and pages about the Debrecen Medical School don’t mention Semmelweis. If I’m wrong about that I may be wrong about this as well, but my recollection is that they spelled and pronounced Semmelweis’s name as in German (both the first S and the final s). The Hungarian version of the Wikiparticle about him calls him Semmelweis Ignác Fülöp: the given names could hardly look more Hungarian, but the family name is anything but.

  118. I’m guessing they just know that in German names they apply different values, just as Russians know that in classical (e.g., Гомер, Медея) and German or German-origin (e.g. Мандельштам) names consonants aren’t palatalized that would be in echt-Russian names.

  119. Athel Cornish-Bowden says:

    I went to a meeting in Debrecen in 1985

    An anecdote about that that the Russian speakers here might like to comment on.

    There were quite a few Russians at that meeting, and they spent as much time as they could in downtown Debrecen. Many of us wondered why, as to our western eyes downtown Debrecen wasn’t very interesting, so someone asked one of them, who said “Debrecen may not seem like much to you, but to us it’s Paris”. The point was that they could buy things like shoes that were not easily available in the USSR.

    I saw a bit of that much more recently at a Beilstein meeting in Rüdesheim. There were two young women from Kazakhstan. I think they had Kazakh passports, but they were clearly Russian in other respects. They attended the first half of the first session and then were never seen again. We assumed that they regarded their grants to attend the meeting (which I think came from the Beilstein Institute, not from the government of Kazakhstan) as a chance to to do some shopping in Germany.

  120. Stu Clayton says:

    # The University of Debrecen, GE Healthcare, the National Institute for Quality- and Organizational Development in Healthcare and Medicines (GYEMSZI) and Semmelweis University are all members of Europe’s largest co-operative partnership in healthcare research development and innovation. #

    http://semmelweis.hu/english/2015/01/hungary-at-the-forefront-of-european-healthcare-innovation/?fbclid=IwAR3sEllmNmSWbq5FGW718Yg3YJ-kOJVPKMcTlBWmFFtegstQiLzzR-YVwMY

  121. A Russian journalist told me a similar anecdote.

    In 2004, there was a young Chechen woman recruited by terrorists to become a suicide bomber. A girl from mountains of Chechnya who knew very little about modern life (and so it was easy to brainwash her).

    Anyway, the terrorists left her in Moscow underground with instructions to blow herself up when there are lots of people around.

    But they didn’t take into account psychology.

    You see, Moscow metro stations have boutiques and shops with wonderful, amazing goods the poor girl never saw in her life. So instead of blowing herself up she spent over an hour looking at pretty clothes in boutiques until she was stopped and arrested by police.

  122. David Eddyshaw says:

    I think this was the key concept behind Ostpolitik.

  123. John Cowan says:

    When I was a child, about 40 years ago, it was still usual for children in Germany to address adults, except for their parents / grandparents, with Onkel / Tante, independent of whether they were related or not.

    That was also the case in my house in the 1960s, although not any and all adults, only the friends (not business acquaintances) of my parents or either of them. This may have had something to do with my German mother, to be sure.

  124. Athel Cornish-Bowden says:

    Something I didn’t notice the first time I looked at the Hungarian page was this sentence:

    A kor következetlen anyakönyvezésére jellemző, hogy a tíz testvér vezetéknevét tízféleképpen írták; Semmelweiss, Semmelweis, Semelveisz stb.

    which means, apparently,

    The inconsistent registration of the age reflects the fact that the surname of the ten brothers was spelled in ten ways; Semmelweiss, Semmelweis, Semelveisz, etc.

  125. All three retain correct pronunciation though. First two follow German spelling and the third Hungarian, but all three are pronounced as Shemmel-vise.

  126. Stu Clayton says:

    Pronounced as “Shemmel-vise” ? Then in what sense do the first two spellings “follow German spelling” ? They would have to start with “Sch-” if the initial sound is that of English “sh”.

    Most of the head-scratching above about how his name is pronounced are about the final consonantal sound. So I’m surprised at “Shemmel-vise”. I hope you can issue a certificate certifying the certainty of that, I need a new insider pronunciation to awe people at medical cocktail parties.

  127. Athel Cornish-Bowden says:

    It’s a pity that they reduced the other seven ways to “etc.”. I’d expect to see something like Szemelvajsz in there.

  128. @Jonathan Does anyone know of some theoretical approaches to semantics other than Natural Semantic Metalanguage?

    I was hoping to see some citations to Linguistic research in response. My impression on first studying Philosophy of Language (mid-1970’s) was that all the hard work was going on in Philosophy depts — of course they would say that; but we had some guest lectures from the Linguistics dept, who didn’t disagree. Seems strange: as if the only part of Language that Linguistics doesn’t study is what Language is for. NSM seems barely any advance on what Linguistics had to offer back then.

    To answer the q, I would look at ‘Montague Grammar’, which despite its name is not really about grammar, but how semantics is expressed in grammar. There’s a wiki, but go from that to the long piece by Barbara Partee. Of course there’s also Gricean Pragmatics, Kripke semantics, and sundry abstruse topics considered by Quine. All of these are research-in-progress, very much not case-closed.

  129. @Bathrobe No, it’s Wikipedia. What do you expect?

    It’s not even Wikipedia; it’s a stub, which means the wiki-police don’t regard it as adequate. I expect that if you give me a citation it’s going to inform the discussion.

    I expect at least definition of Specificity.

    “specificity is a semantic feature of noun phrases ” strikes me as starting straight into the sort of muddle Pullum would fulminate against: a noun phrase is a morphosyntactic category, whereas most of what the piece talks is about semantics (of English — which clearly doesn’t have morphosyntactic marking of specificity). If Samoan has morphosyntactic marking of Specificity, as the wiki claims, it would be nice to see examples of what distinctions it marks.

    As it is, all of the examples the wiki gives (in English) are easily explained using the logic tools of quantification and relative scope — that is, the tools already honed to account for Definiteness, per the Stamford article. In some cases it’ll need scoping of the speakers beliefs/presuppositions, including those relating to the speakers suppositions about hearers’ knowledge. All covered by Grice and/or Kripke.

    Contra the wiki, I see no evidence to distinguish Specificity vs Definiteness — that is, unless there’s a language that morphosyntactically marks both categories.

  130. David Eddyshaw says:

    “Definite” = “already established in the discourse”; specific as opposed to generic. They interact, certainly, but are quite distinct. I don’t think Definiteness as a philosophical concept really corresponds to its normal use in linguistics.

    The marking of definiteness, in particular, is pretty language-specific; it need not involve articles and may only happen with particular syntactic roles, like verb objects.

    Montague semantics addresses a quite different area of meaning from NSM; I would myself say it is even more vulnerable to Wittgensteinian objections, though.

  131. Thanks David, but I think you’re falling into the same muddle of semantic vs morphosyntactic: Definite = marked by ‘the’ vs ‘a’ aot in English definite vs indefinite article. So it’s not a morphosyntactic category applicable in Russian. Yes (definite) Descriptions as a philosphical/logical concept is different, and arguably would have had different terminology if Russell hadn’t been an English speaker.

    English “the” is not restricted to “already established in the discourse” (that’s a semantic observation). I can continue a discourse with “the picture hanging on the wall is askew” without there having been previous mention of pictures or walls; it might be deixis: we’ve been discussing earthquakes; we’re in a room with a picture hanging on a wall.

    Specific vs Generic is not a distinction drawn in that wiki; that alleges it’s to do with uniqueness. Do you have a reference? “The four horseman of the apocolypse” is that Specific despite there being four? “The quality of mercy is not strained as water through a seive.” Despite the “the”, is that Specific or Generic?

  132. David Eddyshaw says:

    As you say, English “the” doesn’t invariably mark “established in discourse”: that is an instance of the very thing I was alluding to, that the marking of “definiteness” is very language-specific.

    In Kusaal, for example, the “definite article” la isn’t used with generic reference (as it can be in formal registers of English), nor with “assumed general background knowledge”: so no definite articles in e.g.

    Winnig tansid ne.
    “[The] sun is shining.” (specific, unique; but not previously mentioned in this conversation.)

    Tumtum pu gaadi o zugdaana.
    “[The] servant does not surpass his master.” (proverb: generic)

    “Definite” articles are named after their prototypical use, not their invariable signification.

    Moreover, “definiteness” in this sense is not always expressed by articles at all. Many languages (Turkish, for example) use accusative case marking only for “definite” objects; Finnish distinguishes definite from indefinite objects by case.

  133. Finnish distinguishes definite from indefinite objects by case.

    It’s not as simple as that.

    Partitiivi indicates …
    http://jkorpela.fi/finnish-cases.html

    Its use is also described in more detail in Finnish: A Comprehensive Grammar (Routledge Comprehensive Grammars).

    Library Genesis and sanet.st, eg, have it.

  134. David Eddyshaw says:

    “Specific” certainly does not entail uniqueness. I have three very specific biscuits in front of me at this precise point.

    Whether abstracts are specific or generic varies along at least two dimensions: countable abstracts, like “question”, behave much like countable concrete nouns; whether uncountable abstracts like “mercy’ or “linguistics” are specific or generic depends on how Platonic you’re feeling (though like concrete mass nouns, they can be constrained to specific senses: “the linguistics practised by Chomsky and his disciples.”) Different languages assimilate them to different types: English “glory” is like “water”; French “gloire” is like “France.”

  135. David Eddyshaw says:

    Thanks, Juha: I was hoping an actual Finn would chime in.

  136. onko totta, että valkoviiniä juova kärsii vähemmän seuraavana päivänä kuin punaviiniä juova?

    is it true (prt) that a person drinking white wine (prt) suffers less the following day than one drinking red wine (prt) ?

    https://en.wiktionary.org/wiki/tosi#Finnish

  137. David Eddyshaw says:

    Any actual Turks out there …?

  138. +
    Hauskaa on, että esimerkiksi housuhameet ovat palanneet.
    It’s fun that. eg, culottes have made a comeback.

    https://en.wiktionary.org/wiki/housuhame

    The word possibly originally meant “wasteful, diminishing, useless” (in which case, compare hupa for semantic shift), in which case it might be related to haaska and derived from haja-.
    https://en.wiktionary.org/wiki/hauska

  139. +
    Luin kirjaa eilen.
    I was reading a/the book yesterday.

    Luin kirjan eilen.
    I read a/the book yesterday (and finished it).

  140. Stu Clayton says:

    housuhameet = “culottes”, sanskulotit = “sans-culottes”. What a demanding language !

  141. David Eddyshaw says:

    Just to complicate the issue further (and why not?), yet a third dimension involved is the distinction between referential and non-referential (of particular significance to me because it affects the form of dependent nouns within Kusaal NPs.)

    I think that non-referentials have to be generic, but that generics can be referential. However, if I think about it for too long I get migraine. Even CGEL seems to struggle with some of these issues. It’s also easy to mistake what are really just questions of how you choose to label things for substantive issues, so that you end up missing the fact that you’re really only arguing about definitions.

  142. Which anglophones still address an aunt or uncle as “Aunt”/”Uncle” tout court, as opposed to “Aunt Jane”/”Uncle John”? I’m guessing the kind of person who calls their father “Father”. It sounds Jane Austen to me, but I’ve heard it from Indian anglophones (where, significantly, it would be indistinguishable from the more general “aunt”/”uncle” of respect if you didn’t know the kinship of the people involved).

  143. Stu Clayton says:

    It’s also easy to mistake what are really just questions of how you choose to label things for substantive issues, so that you end up missing the fact that you’re really only arguing about definitions.

    “Only” arguing about definitions ? That’s one way to put it. Here’s another, more neutral way: you’re “arguing” only in the sense that certain words uttered by each, considered “important” by each, are met by different words from the other, which the other considers “important”. There is no back-and-forth using the same words that both consider “important”. When that occurs, its called “disagreement”.

    If for a moment you think of discussion this way, you free yourself from bootless psychological speculation about what the other guy “means”. You’re free of the ontological shackles imposed by the notion of “definition”, which implies there is “something out there” to which it might apply.

    In this brief moment of relative freedom, the idea can be entertained that you could just blow the whole thing off, and instead stick with those who happen to catch those certain words you consider important, and biff them right back at you.

    Birds of a feather flock together for good reason. It is their way of avoiding fruitless discussions with birds of a different feather, which so often lead to painful pecking and ruffling of plumage.

    Peace, brother !

  144. David Eddyshaw says:

    Sorry, Stu: I just don’t understand what you mean.

  145. Stu Clayton says:

    Excellent ! I see I’ve made my point. It’s behaviorism tempered with self-preservation. No need for overt agreement.

  146. Seems strange: as if the only part of Language that Linguistics doesn’t study is what Language is for.

    Not so strange; how many sciences study what that science is “for”? Do astronomers study what astronomy, or the universe, is for? It’s hard enough figuring out how things work.

  147. David Eddyshaw says:

    I’m guessing the kind of person who calls their father “Father”

    My children call me “Father”; I’m just that intimidating.

    Never heard any of them calling an aunt or uncle “Aunt” or “Uncle” tout court though. Less intimidating, I suppose. There’s also the issue that they tend to encounter aunts and uncles in groups, unlike fathers, come to think of it, so there may be questions of disambiguation involved.

  148. David Eddyshaw says:

    Those who ask the question “What is language for?” very often seem to assume without much discussion that it is basically for communicating by means of true propositions, and that other functions of language are essentially parasitic on this core: this assumption then obscures their entire view of how language actually works.

  149. Good point.

  150. I sometimes call my aunts “Aunty” and my uncles “Uncle”. (I would never call anyone “Aunt”.)

  151. how many sciences study what that science is “for”? Do astronomers study what astronomy, or the universe, is for?

    False analogue. I didn’t say ‘[study] what Linguistics is for’ I said “what Language is for”. And certainly Sociologists study what Religion is for, or what Bureaucracy is for. Language is just as much a human construct; planetary motion not so much.

    Pace David’s presumption: I made no presumption that Language is for communicating; one of its purposes is making true propositions; another (thanks to the arbatriness of the sign) is to make false propositions; another is to make Speech Acts a la Austen.

    It’s hard enough figuring out how things work.

    Linguisticians are injecting semantics into their analysis all the time. I find David E’s explanations a continuing muddle of morphosyntax with semantics, I suspect it’d be easier to figure out how things work if those two were kept separate; which means at least being able to winnow the morphing from the what-for.

  152. David Eddyshaw says:

    Ah. You believe an an empyrean realm where meaning can be grasped in its purity without the sullying interposition of language. A mystic. I respect that. Inevitably, I will never understand it, however …

  153. Athel Cornish-Bowden says:

    Seems strange: as if the only part of Language that Linguistics doesn’t study is what Language is for.

    L-Hat has already commented on this, but let me put my oar in. Extremely few biologists are interested in Life. You can search the 20000 or so pages the Journal of Biological Chemistry published last year with out finding a discussion of what life is or what it is for.

  154. Athel Cornish-Bowden says:

    Pete Olcott, one of the resident nutters at sci.lang (even one of the other nutters think he is a nutter) is obsessed with the idea of semantic atoms and the representation of Truth. I take your semantic primitives to be closely related to his atoms. He has been quiet lately, but I expect he’ll be back.

  155. AJP Crown says:

    Wooster called his aunts tout court “Aunt”. The only person I can think of who was addressed as “uncle” is Guy Crouchback by the younger Halbardiers. So maybe it was popular until the 1950s. Evelyn Waugh’s family called his brother Alec Uncle Sex though not to his face, I think.

  156. John Cowan says:

    Extremely few biologists are interested in Life.

    Peter Medawar, we are told, silenced a whole room of biologists debating the subject thus: “Gentlemen, everyone in this room knows the difference between a live horse and a dead horse. Pray, therefore, let us cease flogging the latter.”

    specificity and definiteness

    Let me recommend Haspelmath’s capitalization convention: comparative concepts, which are invented by us for typological purposes, are in lower case; descriptive categories, which are specific to languages and discovered empirically within them, are capitalized. Thus in English the present tense is expressed by the Present Progressive, and although English has no Dative, it definitely expresses the dative by word order or the Preposition to.

    For my purposes, at least, definiteness and specificity are comparative concepts. A reference is definite if I think you know who the referent is, and indefinite if I have no such view. The Definite Article the may be used in either definite or indefinite references. A specific reference is one whose referent you can only determine with surety by asking me which one you mean. English marks definiteness with its Articles and uses a certain to mark specificity without definiteness: I know which referent I have in mind, but I don’t suppose that you do. Turkish and Persian mark specificity morphologically, and Persian marks indefiniteness with its Indefinite Article.

  157. You believe an an empyrean realm where meaning can be grasped in its purity without the sullying interposition of language.

    Yes, I suspected this when AntC wrote “I don’t see that any theory of semantics which attaches meaning purely to appearances of words is going to be adequate”. The reference to “appearances” is telling. AntC appears to see language as nothing more than shadows on the wall of the cave.

    one of its purposes is making true propositions

    The Stanford Encyclopedia of Philosophy has A LOT to say about Propositions.

    In the introduction it notes as follows:

    If David Lewis (1986, p. 54) is right in saying that “the conception we associate with the word ‘proposition’ may be something of a jumble of conflicting desiderata,” then it will be impossible to capture our conception in a consistent definition.

    The best way to proceed, when dealing with quasi-technical words like ‘proposition’, may be to stipulate a definition and proceed with caution, making sure not to close off any substantive issues by definitional fiat.

    Suffice it to say that quite a lot of discussion of this “quasi-technical word” relates to the properties of natural language, for instance, the Substitution Problem that deals with verbs in natural language.

    I don’t think any “linguisticians” agree on the cutoff between morphosyntax and semantics. AntC obviously has his view, that semantics is separate from “the appearances of words”, and that semantics is properly concerned with the “truthhood” of Propositions. All well and good, but this seems to be a very narrowly circumscribed, indeed, dogmatic view of the function of language.

    @JC

    Nicely stated.

    English marks definiteness with its Articles and uses a certain to mark specificity without definiteness

    And my point, made long, long ago, was that colloquial English can do the same thing by using this.

  158. David Eddyshaw says:

    The prototypical (OK, OK, I like that word, OK?) context for specific indefinites is in introducing a new discourse topic; Kusaal can do that just with an NP without the article (there’s no “indefinite article.”) Thus, beginning a story:

    Dau n da bɛ.
    (Man:SG CATENATOR PAST exist.)
    “Once there was a man …”

    where the catenator n follows a focused subject.
    But more often, Kusaal uses the (specific) specific-indefinite pronominal sɔ’~si’a~sieba “some, a certain, another, a different” after the head noun:

    Dau sɔ’ n da bɛ.
    (Man-INDEF.ANIMATE CATENATOR PAST exist.)

    [dau sɔ’ is in fact a compound, though this is disguised by the standard orthography, which omits tone marking and has inconsistent word division conventions.]

    Or in Bathrobics: “There was this man …”, with “this” having exactly the specific-indefinite sense of sɔ’.

  159. January First-of-May says:

    or you could say German has lost “GO”

    Mine has. I have no single word for “locomote”.

    It’s probably a common situation that a language doesn’t actually have a sufficiently generic word – and likely an even more common one when the only options of sufficient genericity (generality?) are far too generic.

    (For example, there was no Russian word for “sibling”, so when Russian scientists had to talk about siblings, they borrowed the English word.)

    It’s one of my favourite examples of cultural conservatism: in the West we wear trousers because the barbarians who overran the western provinces of the the Roman Empire did.

    Having known from childhood the name пифагоровы штаны “trousers of Pythagoras” for Euclid’s windmill, and having seen many assorted comments regarding how Pythagoras was an Ancient Greek and thus probably did not wear any trousers, I was quite surprised when I found out that the Roman historian Aelian claimed that Pythagoras did wear trousers (apparently because he was fascinated by then-eastern barbarians).

    Mind you, Pythagoras of Samos was already a semi-legendary figure by Aelian’s time; by now, supposedly, there’s even less evidence for a historical Pythagoras of Samos than for a historical Jesus of Nazareth (though, weirdly, I have not heard of anyone seriously proposing that Pythagoras never existed).

    Also, Vladimir Plungian is a cool name. It just is.

    I agree!

    Plungian on LH previously (linking to an even older mention).

    Russian has seven basic colour words, but not all languages have exactly that number.

    Russian has seven basic colors of the rainbow, which is indeed (said to be) quite unusual (English had to shoehorn a non-basic seventh color in; I’ve heard that some other languages didn’t, and just have six or even five).

    However, as had been correctly pointed out later, there’s at least five non-rainbow basic colors in Russian: white, black, grey, pink, and brown.

    One minor example of this is that Harry was forced to call his uncle’s sister ‘Aunt Marge’.

    In Russian, the point of this was missed in a different direction: the word for “aunt”, тётя, is a (mildly) disrespectful term of reference to (adult female) strangers (especially from children).
    So from a Russian perspective, of course Harry would have called Marge тётя, because she’s an adult female who isn’t directly related to him, and he doesn’t like her; indeed, as I understand it, in a Russian context, he could hardly have called her by any other term unless he was actively being polite (or, I suppose, actively being impolite).

    [EDIT: the term had in fact been so bleached in Russian that a briefly-popular madlib generator included the surprisingly-unremarkable-sounding sentence Моя мама очень злая тётя – literally “My mom is a very evil aunt”.]

    Even for legitimate aunts, things can get tricky. As a child I commonly visited (and as an adult was even more commonly visited by) my Aunt Tanya – who was actually my двоюродная тётя “second-order aunt”… or, in English kinship terms, “first cousin once removed”.
    (I did have an actual aunt – Aunt Leah, my father’s younger brother’s wife – but hardly ever met her, because she lived in Israel. She and my uncle have since divorced, so I have no idea where she lives now, and I’m not even sure if she still counts as my aunt.)

    […Fun and weird fact I just realized: “first cousin once removed” is a symmetric term of kinship – if A is B’s first cousin once removed, then B is A’s first cousin once removed, and vice versa.
    Are there any non-English examples of symmetric kinship terms referring to relatives outside the same generation?]

  160. Моя мама очень злая тётя

    Which I want to sing to the tune of “У самовара я и моя Маша.”

  161. Sorry, doesn’t scan. “У самовара…” is iamb so you need an extra syllable between Моя and мама. Something like “Моя-то мама…” or “Моя же мама…” or, you know, there is a universal Russian discourse particle, but I am a bit tired of this aspect of my native tongue and let’s stick with these two.

  162. For example, there was no Russian word for “sibling”, so when Russian scientists had to talk about siblings, they borrowed the English word.

    And the English word “sibling” was dug out of archaic English by anthropologists, since modern English didn’t have a word for it.

  163. And very strange it is too, since it seems just as obvious a concept as “parents.”

  164. Trond Engen says:

    Scandinavian languages have søsken/syskon “siblings”, a plurale tantum with a fossilized plural ending. At least Norwegian has reinterpreted it as a neuter plural to produce an identical singular (et) søsken.

    The second level is søskenbarn lit. “children of siblings” -> “cousins”. This also started as a plural but has been reinterpreted so that it can be used with reference to the children instead pf the parents. Vi er søskenbarn “We are sibling-children (-> cousins)” ~ Jeg har sju søskenbarn; “I have seven cousins”.

    The third level is tremenning(er) “second cousin(s)”. As far as I know, this always had a singular form. Then there are firmenning(er), femmenning(er), etc, ad infinitum.

    A second cousin once removed used to be a halvtredjemenning, and so on, mutatis mutandis, for higher orders of cusinity, but this is now archaic-to-obsolete.

  165. David Eddyshaw says:

    Kusaal has no word completely equivalent to “sibling”: the nearest equivalents are the compounds mabiig “child by same mother” and ba’abiig “child by same father”, though in the context of traditional family structures these are respectively more-or-less equivalent to “full sibling” and “half-sibling” respectively.

    Interestingly, the Bible translation picks ba’abiis for the brothers-and-sisters-in-Christ “brethren” sense.

  166. David Marjanović says:

    German has Geschwister (plural and rare backformed neuter singular), which – in flagrant violation of the patriarchy – must have been the collective of “sister” once.

    (Gebrüder exists, but is quite archaic and sometimes replaces the ordinary plural Brüder in Gebrüder Grimm and the name of at least one corporation).

    And very strange it is too, since it seems just as obvious a concept as “parents.”

    But then, parents is a loan, and the German version Eltern is almost transparently die Älteren, “the elders”/”the older ones”.

  167. Greg Pringle says:

    Japanese has 兄弟 kyōdai (elder brother + younger brother) referring to all brothers and sisters. Japanese 姉妹 shimai (elder sister + younger sister) refers only to female siblings.

    Chinese 兄弟 xiōngdì and 姐妹 jiěmèi seem to mainly refer to brothers and sisters respectively. For ‘brothers and sisters’ it has 兄弟姐妹 xiōngdì jiěmèi.

    (Online sources say otherwise, so perhaps there is some indeterminacy in broader usage. I think the above represents ordinary usage.)

    But I can’t think of a term for a sibling in the singular.

  168. David Eddyshaw says:

    Kusaal also has no word equivalent to “parents” (though of course there’s nothing to stop you saying “father and mother”); the nearest equivalent is du’adib, transparently the plural of the agent noun du’ad of the verb du’a, which in proper unisex fashion means both “bear” and “beget.” However, du’adib actually has a broader reference, and applies to all (senior) blood relatives (rather like French parent, now I think of it, which come to that is pretty much parallel in etymological terms to du’ad as well.)

    The proper way to express “parents” in Indoeuropean is of course to use the dual of “father.” Or “mother” if you’re a Woke Indoeuropean.

    Did French parent once mean “parent”, by the way? Some Hatter will know … [indefinite, yet specific]

  169. David Marjanović says:

    the dual of “father.” Or “mother” if you’re a Woke Indoeuropean.

    Both of these are actually attested in Sanskrit.

  170. John Cowan says:

    Interestingly, the Bible translation picks ba’abiis for the brothers-and-sisters-in-Christ “brethren” sense.

    Inevitably so. God is our common Father, but that does not entail the existence of a common Mother, at least not a Christian one. Indeed, I shouldn’t wonder if the objection of Nestorius to the title Theotokos ‘Mother of God’ was that he suspected his Alexandrian opponents of crypto-Isisism.

    “the elders”/”the older ones”

    The OE word was ealdor (m), but whether it was used in the plural, I have no idea.

  171. David Eddyshaw says:

    Inevitably so

    True enough (as occurred to me after the editing window closed …)

    Still, it’s interesting to see how the metaphor becomes rather more obviously metaphorical in translation to a culture with different ideas about what typical families look like.

  172. Lars Mathiesen says:

    Parents are padres in Spanish, of course. (Not a woke language, but we knew that). I see L patres attested in that sense once in an inscription, but mostly meaning ‘ancestors.’

    Danish has søskende for ‘siblings’ — the plural meaning was mandatory 50 years ago, but the lack of a singular was not really a problem; even if you wanted to be nonspecific, you’d say “we are two siblings” instead of “I only have one sibling.” I think YPNAD can use it as a singular. *brandishes cane*

    For some reason the ending has acquired the same spelling as a present participle, which confuses the issue. The latter shows no number distinctions in modern Danish (and cannot take the definite marker) so there is no way to make an overtly singular form. (The participle had original *-nT which regularly became /ɲː/ spelled {nd}, but ON systkin looks like it had a short nasal so there is no excuse).

    Scandinavian also has forældre for ‘parents’ — at least Danish, I think it’s basically the same in Sw and Nw — again it used to be plural only, because an individual parent is a mother or a father and the need for a neutral term wasn’t felt. Now it is and we say forælder in Denmark, to the horror of some.

  173. David Marjanović says:

    The singular Elter occurs only in Mendelian genetics. It’s masculine of course.

    The normal singular, however, is Elternteil – also masculine, but that’s the fault of Teil “part”, nothing we can do about it.

  174. Athel Cornish-Bowden says:

    Did French parent once mean “parent”, by the way? Some Hatter will know … [indefinite, yet specific]

    In the French I hear every day parent still means parent in appropriate contexts. If someone says mes parents you can usually be sure which relatives they mean. But maybe you’re just asking about the singular.

  175. Trond Engen says:

    I meant to write about foreldre and the singular forelder, backformed on the Dano-Norwegian pattern used in Bokmål for agent nouns (en bærer ~ to bærere). I half-expect the verb forelde “make obsolete, expired” to take on a meaning “to parent”.

  176. Trond Engen says:

    Also, Lars is right and I had forgotten about the word søsken.

  177. David Marjanović says:

    forelde “make obsolete, expired”

    Ha, the causative to veralten “become obsolete”!

    Veraltet is “obsolete”.

  178. Lars Mathiesen says:

    agent nouns — in Danish at least, I think a more obvious model for the singulative of forældre is fader/fædre, moder/mødre (and non-kinship terms like alder/aldre that don’t umlaut because a-stem).

  179. Trond Engen says:

    Yes, perhaps. Agent nouns are just a part of what goes into that conjugation. Forelder is longer and more like a verbal noun than the other examples, but that doesn’t mean identification with verbal nouns.

    The causative form forelde is a little artificial. The meaning “expire” is expressed by the passive foreldes. Also common are the participle forelda/et “expired” and the verbal noun foreldelse “expiration”.

  180. Trond: A second cousin once removed used to be a halvtredjemenning, and so on, mutatis mutandis, for higher orders of cusinity, but this is now archaic-to-obsolete.

    What might cause it to become obsolete when they’re still around?

    How do I say “first cousin once removed”? I have plenty of them, both older and younger, and I usually resort to søskinbarn, fetter or kusine (I’ve found the vagueness causes some confusion).

  181. Trond Engen says:

    They’re archaic-to-obsolete because very few know them and nobody uses them anymore, except maybe in a historical context. But they are good words that ought to be resurrected.

    A first cousin once removed is a niece or nephew or uncle or aunt. We don’t have a common term for all of those, but I guess you could coin halvannenmenning.

  182. John Cowan says:

    Indeed, there has been a breakdown in trensmission of these terms in English too. Though we still have terms like second cousin, there is no agreement on what they mean.

    The tradiitonal view is that second cousins are the children of first cousins, and so on: your nth cousin is a person whose common ancestor is your ancestor in the n+1th degree (your zeroth cousin would be your sibling, except that the term is not used) Likewise, your nth cousin k times removed is a descendant or ancestor in the kth degree of your nth cousin.

    But now there are people who simplify the system, and for whom their second cousins include their first cousins once removed, and so on. These people have typically never heard “k times removed”, or if they have they don’t know what it means.

  183. I’ve always known there were second (etc.) cousins and cousins once (etc.) removed, but at no point in my life could I have told you what exactly they were, and I suspect I’m typical of most Americans who are not waist-deep in genealogical research.

  184. Stu Clayton says:

    It’s a wise child that knows its (n, k) cousin.

  185. A first cousin once removed is a niece or nephew or uncle or aunt.

    A first cousin once removed is either a) the child of your first cousin or b) your parent’s first cousin. A first cousin twice removed is a) the grandchild of your first cousin or b) your grandparent’s first cousin.

    The traditional view is that second cousins are the children of first cousins
    Yes. They are the same generation.

    In my case, it hasn’t got anything to do with genealogical research only that my grandmother had six siblings who were orphaned at a young age and their families have kept in pretty close touch (even in Australia & Canada).

  186. Trond Engen says:

    Ah, my bad. I was supposed to know that. And it’s obviously more meaningful.

    New attempt: A first cousin once removed is a halvtredjemenning. Halvtredje is the justified and ancient way to say 2.5 (as still in Danish counting of half twenties).

  187. Stu Clayton says:

    Halvtredje is the justified and ancient way to say 2.5

    So this is like the German halb drei to mean the time 2:30 (am or pm) ?

    Although Germany uses the 24-hour numbering system, the corresponding 24-hour nomenclature in speech is more or less optional. Instead of “zweiundzwanzig Uhr” you can say “zehn Uhr”, as long as 22:00 is close enough to the time at which “zehn Uhr” is said that no confusion can arise.

    One thing that is only rarely done, if at all, is to use that “halb …” construction with times greater than 12. “halb zweiundzwanzig Uhr” sounds a bit strange. I would never say it, since I have a reputation to lose.

  188. Trond Engen says:

    Yes. Halv tre up here.

    Though that’s odd too. Halvtredje is formally an ordinal number — naturally, since it’s halfway on the third count. I don’t know how and when the form halv tre became established for half hours on the clock.

  189. Stu Clayton says:

    Not only that: what about “dreiviertel vier”, “viertel sieben” and so on.

  190. AJP Crown has the traditional definitions of removed cousins right. I suspect that was what John Cowan meant as well, but he got it muddled, resulting in a nonreciprocal relation.

    To be very precise, nth cousins k times removed are two people who share a common ancestor (or ancestor couple) n + 1 generations back (for one of them) and n + k + 1 generations back (for the other).

    Separately:

    @Stu Clayton: I have heard X:30 as “half X” once or twice from British speakers, although I could not tell you what particular regional dialects those people spoke natively.

    If I heard “dreiviertel sieben” in German, I would interpret it as “dreiviertel nach” rather than “dreiviertel vor.” I’m not sure how standardized that is, but “viertel sieben” is definitely vor.

  191. Stu Clayton says:

    but he got it muddled, resulting in a nonreciprocal relation.

    cousins … n + 1 generations back (for one of them) and n + k + 1 generations back (for the other).

    That (n, k) statement is reciprocal only for cousins. It wouldn’t be reciprocal for aunts, for instance, because the aunt relationship is not verbally symmetric (that’s not a particularly high-class way of phrasing the matter, but I can’t think of a better offhand that doesn’t require several sentences of exposition).

    Are aunts, uncles etc not supposed to be specified when talking (n, k) ? I’ve never understood that whole business, I suspect because I think I’ve heard things like “aunt twice removed”.

  192. AJP Crown says:

    New attempt: A first cousin once removed is a halvtredjemenning.
    Aha, thanks!

    Halvtredje is the justified and ancient way to say 2.5 (as still in Danish counting of half twenties).

    And since it’s 2.5, halvtre is two-thirty (i.e. not three-thirty, as in English) on the clock. This can be fatal when you’re arranging meetings.

  193. Stu Clayton says:

    And since it’s 2.5, halvtre is two-thirty (i.e. not three-thirty, as in English) on the clock.

    Do you mean that “half three” could be confused with “three and a half” ? It’s a good thing we have neither for clock times.

  194. AJP Crown says:

    “aunt twice removed”
    You remember that time your uncle went bankrupt? First they disappeared to the other side of town and then they were never seen again?

  195. Stu Clayton says:

    I guess I was confusing consanguinity and liquidity. Of course blood is thicker than water.

  196. @Stu Clayton: The “cousin” terminology is, of course, only used for n ≥ 1 and k ≥ 0, with the “times removed” omitted for k = 0. Standard for a sibling of a direct ancestor n + 1 generations back (n ≥ 0) is “aunt” or “uncle” preceded by n “great”s. Dialectically, the first “great” may be replaced by “grand,” as is standard with direct ancestors. Bilbo was spluttering a bit when he said, “I had a great-great-granduncle once, Bullroarer Took…,” and as well as sounding rural, he gets the number of generations wrong. (However, it is unclear whether the mistake was actually Tolkien’s, when, years later, he was preparing the family trees of the Shire gentry for The Return of the King.)

  197. AJP Crown says:

    Do you mean that “half three” could be confused with “three and a half” ? It’s a good thing we have neither for clock times.

    I don’t kno, maybe it’s a British thing:

    “What’s the time?”
    “It’s half three.”

    But now there’s this expression “See you in ten,” with the minutes missing. That only reached my desert island recently, along with “Mine” for “My place” or “My house” – this latter expression I got from Drama (= BBC cop shows) off the telly.

    Brett: “I had a great-great-granduncle once,

    I knew my great-great aunt pretty well, my great-grandfather’s youngest sister. She died when I was about seventeen. I agree that 3x great is pushing it but it must happen quite often nowadays (with longer lives).

  198. Lars Mathiesen says:

    Danish has a traditional system with terms like grandonkel, grandfætter and grandnevø, which according to the encyclopaedia would be the uncle/cousin of one of your parents, and the son of your nephew or niece. But outside the encyclopaedia, confusion reigns supreme.

    (One problem is that there is no opposite term to grandfætter).

    We do have the direct line of ascent/descent sorted, though: oldemor/far (great grandparents), tipoldemor/far, tiptipoldemor/far and so on, oldebarn etc. in the other direction.

  199. Stu Clayton says:

    along with “Mine” for “My place” or “My house” – this latter expression I got from Drama (= BBC cop shows) off the telly.

    As in “mine or yours” instead of “my place or yours” for trysting locations ? Is that the New Banality ? Doesn’t sound like either party is particularly interested in the choice of trist location.

  200. Athel Cornish-Bowden says:

    I think that there has always been confusion about “removed”, and it’s not going to change for the better in the future. However, I’m not convinced that there is much confusion about second, third, etc. I have a third cousin that I have known well since we were children; I used to know her two sisters well as well, but I haven’t seen them for a long time. On account of a cousin marriage they’re also my sixth cousins, but I wouldn’t have known that if there hadn’t been any genealogical research. On the other hand no genealogical research was necessary to know they were third cousins, because the families have always been in touch.

  201. John Cowan says:

    but he got it muddled, resulting in a nonreciprocal relation.

    I thihk the definitions are equivalent, but yours is more elegant because it doesn’t require that the linking relatives actually exist. If Alice (whose daughter is Anna) and Bob are (1,0)cousins, then Bob is Anna’s (1,1)cousin, because she is the 1st-degree descendant of Bob’s (1,0)cousin. By the same token, Anna is Bob’s (1,1)cousin because he is the 1st degree ancestor of Anna’s second cousin Bill (son of Bob), but that definition breaks down if Bob is childless.

    aunt twice removed

    ChiCha : So, remind me again how you’re related to Pacha?

    Yzma : Why, I’m his third cousin’s brother’s wife’s step-niece’s great aunt. Twice removed.

  202. Stu Clayton says:

    It’s slightly odd that statements like “seals are more closely related to weasels than to whales” (or things to that effect, said here recently) trip off the tongues of experts, but the analogous “removed” relationship for humans remains shrouded in semantic and genetic (“it’s a wise child …”) uncertainty.

    Maybe it’s because species and genera are being compared in the former case. There might be an Eddy Seal more closely related to Mary Whale than to Willy Weasel, but nobody cares. Any significance is swallowed up by statistical generalization, except for the desire to say something Significant because General.

    Since seals don’t have public marriage ceremonies, and the locations of the registry offices are unknown, it’s not surprising.

  203. January First-of-May says:

    I agree that 3x great is pushing it but it must happen quite often nowadays (with longer lives).

    Not even that long; just needs shortish generations on one side, and a largish gap between siblings on the other.

    Even in my own family, one of my grandmother’s brothers already has teenage great-grandchildren (who are thus my second cousins once removed), while another of her brothers (and thus their great-granduncle) is only 78 (born 1941) and could easily live long enough to see them have their own kids.

  204. Sometimes, kin terms get stuck at a certain point, if children start calling someone the same thing as their parents. This is probably particularly common for relatives with whom one has an especially close personal relationship. My grandfather called his favorite uncle’s wife “Aunt Fanny,” and so did my father, and so did I, even though she was technically my great-great-aunt. She lived past one hundred (outliving my grandfather, in fact), but my kids never got to meet her. However, her name lives on in family recipes, like Aunt Fanny’s Strudel.

    From the same generation: My grandmother called her father “Pop,” so everyone else did as well. It may have helped that he was not that fond of his given name, “Nathan.” More recently, my kids mostly call my Uncle Dan “Uncle Dan,” because I tend to forget to append a “great” when talking about him.

  205. John Cowan says:

    Dreiviertel at the Hat in 2011 and in 2016.

  206. January First-of-May says:

    Sometimes, kin terms get stuck at a certain point, if children start calling someone the same thing as their parents. This is probably particularly common for relatives with whom one has an especially close personal relationship.

    This can happen in the opposite situation as well – if I mainly know of (e.g.) Uncle Aaron from what my older relatives told me, I might end up thinking of him as Uncle Aaron even if he’s only their uncle and not actually my uncle.
    (I suspect that this is what happened with your Uncle Dan.)

    Ironically enough, this means I have two Aunt Tanyas, neither of whom is technically my aunt – one is actually my first cousin once removed (and often visits me), while the other is actually my great-aunt (to the best of my knowledge, I haven’t met her yet).

  207. David Marjanović says:

    One thing that is only rarely done, if at all, is to use that “halb …” construction with times greater than 12. “halb zweiundzwanzig Uhr” sounds a bit strange. I would never say it, since I have a reputation to lose.

    The systems are never mixed; it’s halb zehn or einundzwanzig Uhr dreißig.

    In stark contrast, in French you can encounter phenomena like midi trente.

    I have heard X:30 as “half X” once or twice from British speakers, although I could not tell you what particular regional dialects those people spoke natively.

    Supposedly it’s quite common in England; definitely not outside the UK.

    blood is thicker than water

    Steirerblut ist kein Himbeersaft “Styrian blood is no raspberry juice”.

    (Though that’s about how manful those men are, not about their views on family. I provide it for entertainment purposes.)

    Maybe it’s because species and genera are being compared in the former case.

    It does help that we operate at a resolution where the tree only branches and the branches never fuse. But the real difference is that we* actually visually imagine the tree when we say such things. Imagining your family tree and then zooming in on it doesn’t seem to be common.

    * No idea how aphantasia might interfere.

  208. @David Marjanović: My high school German teacher posted a different German Sprichwort each week, which was worth a few extra credit points to translate. One week, he put up “Blut ist dicker als Wasser,” which is pretty obvious, since four of the five words are direct cognates to those in the English expression, and als is also easy. However, one of my classmates had a bit of trouble with it. His name was Joshua Jungblut, but when asked about his surname, he always said to pronounce it “Youngblood.” I thought that pronunciation was a bit of a silly affectation, but I couldn’t help but guffaw when he, during some down time in class, was trying his damnedest (and rather loudly) to find a reference for what Blut meant.

  209. David Marjanović says:

    …Oh dear.

  210. At some point, your relations become what Russians call седьмая вода на киселе — seventh water on kissel (google it, my spellchecker doesn’t know the word).

  211. AJP Crown says:

    Supposedly it’s quite common in England.
    Half six fix

    their great-granduncle is only 78 (born 1941)
    And so the same age as Bob Dylan, Charlie Watts, Linda McCartney & Bernie Sanders.

  212. @AJP Crown: You know Linda McCartney is long dead, right?

  213. AJP Crown says:

    ‘Long’ in the same sense as ‘a week is a long time in politics’. Tadao Ando and Toyo Ito then, but Linda McCartney seems younger (35ish).

  214. My father grew up with several brothers and one sister. She was of course Sis to everybody. When I was a boy, she was my Aunt Sis.

Speak Your Mind

*