Hunter-gatherer languages

[Gareth3]

Are there any language features that are more common in hunter-gatherer cultures than in farmer cultures? I ask because I'm creating a conlang for a hunter-gatherer culture. I can come up with features of my own that seem appropriate (like all animals being "neuter" gender because they don't do animal breeding), but I'd be interested in any real-world examples.

[Miekko]

Are there any language features that are more common in hunter-gatherer cultures than in farmer cultures? I ask because I'm creating a conlang for a hunter-gatherer culture. I can come up with features of my own that seem appropriate (like all animals being "neuter" gender because they don't do animal breeding), but I'd be interested in any real-world examples.

What's the causal connection between "don't do animal breeding" -> "animals are neuters". My people don't do clock breeding, yet we have feminine gender on clocks.

[krinnen]

I'd guess there wouldn't be a strong pressure to develop a complex counting system. I'm a bit rusty on this, but IIRC there was an amazonian language (Piraha was it?) with terms for 'one' and 'two', or 'few' and 'many' (distinguished only by tone?). Now, you could devise a counting system with terms for 'one', 'two', 'lots' and 'a basket full', for example.

God, I'm rusty

[Drydic]

The short answer to your question is no, there are not.

[gach]

Indeed, there are no grammatical features specially preferred by the languages of hunter-gatherers, or people from any other type of cultures. If you do a bit of research of Australian and Khoisan languages for example, you'll notice that there isn't too much in common with them. It's also worthwhile to look at languages spoken in post industrial cultures and how the cultural changes have affected the languages. You'll definitely find a lot of new vocabulary (loaned or internally coined) but nothing that would make the grammars form a distinct class apart from other languages.

For grammatical gender the key idea is that it's grammatical. The genders might have something to do with the physical world or there might be no semantic logic at all behind the gender assignment. In Yimas for example you have four semantically based classes for male and female humans and culturally important animals and plants. But the language also has seven further noun classes for all other nouns, which are assigned purely based on the shape of the word. You'll find things like "chair", "mountain" and "snake" fall into one class, "axe" and "breast" into another and "knee", "road" and "side of abdomen" into a third.

You might say that such gender assignment tendencies as to how vigorously a language imposes natural masculine/feminine distinction on human nouns or where exactly you draw the line between animate and inanimate gender are culturally driven. But mostly grammatical gender is a morphological phenomenon with little concern on real world semantics.

[Whimemsz]

The short answer to your question is no, there are not.

This, basically.

Besides, why would a hunter gatherer society place any less of an importance on animals than an agricultural society? Hunter gatherers depend on animals to survive too (or at least they believe they do, even in cases where from a practical standpoint they don't). Language and culture can reflect one another, but it's pretty simplistic to think that the culture will correspond to grammatical gender assignments in such a Whorfian way.

I'm a bit rusty on this, but IIRC there was an amazonian language (Piraha was it?) with terms for 'one' and 'two', or 'few' and 'many' (distinguished only by tone?)

No, the whole reason Piraha is famous is that is has no number works, and the Piraha people are unable to count, learn math, etc. There are a large number of languages with only a handful of number words, including many in Amazonia, and your basic point is correct, that a hunter gatherer society will more than likely have several number words, but not nearly as many as an agricultural society.

[Gareth3]

Thanks for the answers.

[finlay]

The short answer to your question is no, there are not.

This, basically.

Besides, why would a hunter gatherer society place any less of an importance on animals than an agricultural society? Hunter gatherers depend on animals to survive too (or at least they believe they do, even in cases where from a practical standpoint they don't). Language and culture can reflect one another, but it's pretty simplistic to think that the culture will correspond to grammatical gender assignments in such a Whorfian way.

I'm a bit rusty on this, but IIRC there was an amazonian language (Piraha was it?) with terms for 'one' and 'two', or 'few' and 'many' (distinguished only by tone?)

No, the whole reason Piraha is famous is that is has no number works, and the Piraha people are unable to count, learn math, etc. There are a large number of languages with only a handful of number words, including many in Amazonia, and your basic point is correct, that a hunter gatherer society will more than likely have several number words, but not nearly as many as an agricultural society.

Pirahã... well, I've always been skeptical about that. At uni a few of us found a paper by Everett and had a look through, and found that it was very contradictory. In fact an example sentence was used that glossed a word as "two", which sort of refutes their whole point.

Plus I saw a video endorsed by the leaders of the Pirahã council in which they essentially said that Everett and his fellow missionaries were exploitative as fuck, especially when they profited off a film about the tribe without them ever seeing any benefits. They asked them about the teaching maths anecdote but didn't get an answer because it happened in another village, but the guy speaking spoke Portuguese pretty well including words which definitely refer to numbers. They seem to be very pragmatic: they don't do things unless it has obvious benefit to them. In the case of mathematics, it may well be the case that the linguists looked at them too much like experiment subjects and didn't bother to show them why they might need numbers.

On a scientific level, Everett's claims are also unfalsifiable because he's basically the only outsider to learn the language.

[merijn]

I vaguely recall reading a paper (I think written or co-written by Johanna Nichols) that argued that different types of society had different grammatical features. It didn't say anything Wholfian, though, but claimed (and this is all from memory, it could be all different than I remember) that 1) languages spoken by relatively many people evolve quicker than languages that are spoken by only few people and therefore 2) smaller languages had more chance to retain some quirky features.

[Nortaneous]

that sounds like a general case of "urban languages evolve more quickly than rural ones"

[linguoboy]

I vaguely recall reading a paper (I think written or co-written by Johanna Nichols) that argued that different types of society had different grammatical features. It didn't say anything Wholfian, though, but claimed (and this is all from memory, it could be all different than I remember) that 1) languages spoken by relatively many people evolve quicker than languages that are spoken by only few people and therefore 2) smaller languages had more chance to retain some quirky features.

Other linguists have pointed out that it's not only easier to retain quirky features with a small and peripheral community of speakers but to innovate them as well. I remember years ago reading an article to this effect which took as its example pre-occlusion in Cornish, Manx, and one of the Rumantsch varieties (Surmeiran?).

[Whimemsz]

Pirahã... well, I've always been skeptical about that. At uni a few of us found a paper by Everett and had a look through, and found that it was very contradictory. In fact an example sentence was used that glossed a word as "two", which sort of refutes their whole point.

Well, in the numbers/maths/counting case, it's not just Everett who's made this claim. Unlike some of his other claims, in this instance other researchers have found the same thing. (Though apparently some of the other researchers have been more radical here than Everett, in that he's never claimed they are incapable of counting, and that was Peter Gordon instead? Which I didn't remember.)

Plus I saw a video endorsed by the leaders of the Pirahã council in which they essentially said that Everett and his fellow missionaries were exploitative as fuck, especially when they profited off a film about the tribe without them ever seeing any benefits.

Everett is not a missionary anymore (his work with the Piraha led him to abandon his faith -- I don't recall how long it's been, but many years). However it still may well be true that he has exploited them for his own purposes.

On a scientific level, Everett's claims are also unfalsifiable because he's basically the only outsider to learn the language.

Oh I have plenty of doubts about many of Everett's claims. (Anyway, plenty of other linguists have attempted to refute him using his own data. His data are no more unfalsifiable than that presented by any other linguist who has learned a language that no other linguist has.) My impression is that the numbers thing is the least controversial claim he's made, since other researchers have corroborated it and since we know of other cultures with very small and possibly nonexistant numeral systems.


EDIT: Also apparently some other researchers claim that while it's true the Piraha have no number words, they are still able to count and to understand the concepts involved. So.

[Basilius]

I vaguely recall reading a paper (I think written or co-written by Johanna Nichols) that argued that different types of society had different grammatical features. It didn't say anything Wholfian, though, but claimed (and this is all from memory, it could be all different than I remember) that 1) languages spoken by relatively many people evolve quicker than languages that are spoken by only few people and therefore 2) smaller languages had more chance to retain some quirky features.

Other linguists have pointed out that it's not only easier to retain quirky features with a small and peripheral community of speakers but to innovate them as well. I remember years ago reading an article to this effect which took as its example pre-occlusion in Cornish, Manx, and one of the Rumantsch varieties (Surmeiran?).

And, if we define "quirky" as "not occurring in BIG languages"...

[zompist]

It seems to be true in general that hunter-gatherers and garden agriculturalists can get by with minimal number systems. The languages where only a few numbers are reported cluster in Amazonia and Australia. (I only know of two languages which are claimed to have no numbers at all: Pirahã and Yumbri, which is from SE Asia.)

But! I take all that with a grain of salt. If you're actually a field linguist, you should be asking these questions:

1. What are the number words?
2. How do people count out loud?
3. What other ways do they count?

From reading many grammars, it seems many investigators don't get beyond step 1. That is, they look for nice number roots as in Indo-European. Many languages only have one or two number roots, but get up to six by combining them— but grammars or word lists may only have 'one' and 'two'.

Most number systems are based on finger-counting, which is another way early investigators could be misled— a language might have roots only for 1 and 2, but then use words like 'hand' or 'man' to refer to higher numbers. There are cases where the names for numbers are related to finger-counting but aren't fixed.

What arouses my suspicions about all this is within a single language family, some grammars provide a full set of counting numbers, some only supply 1 and 2. I find it easier to believe that some linguists are lazy than that the idea of expressing 3 as [2] [1] is only known by the next tribe over.

Plus, some folks count without saying anything out loud— they just happen not to use number names.

Finally, note that not all hunter-gatherers are satisfied with minimal systems. The Pomo of California could count well into the tens of thousands; this is said to be because they provided clam-shell beads, used as currency, to all the surrounding tribes and thus became experts in counting.

[Radius Solis]

To the best of my understanding, the claim about the Pirahã is not that they are unable to learn to count in principle, but that they find it hard to maintain interest in the topic long enough to get very far with it, in much the same way that most people even in developed, educated countries find it hard to sustain interest in math without a compelling reason like the need to graduate. Even among us, most do not retain very much math ability past school. Even simple things that would be occasionally useful are commonly learned only grudgingly and then forgotten fast; so it is with the Pirahã and counting.

[Dewrad]

I vaguely recall reading a paper (I think written or co-written by Johanna Nichols) that argued that different types of society had different grammatical features. It didn't say anything Wholfian, though, but claimed (and this is all from memory, it could be all different than I remember) that 1) languages spoken by relatively many people evolve quicker than languages that are spoken by only few people and therefore 2) smaller languages had more chance to retain some quirky features.

Other linguists have pointed out that it's not only easier to retain quirky features with a small and peripheral community of speakers but to innovate them as well. I remember years ago reading an article to this effect which took as its example pre-occlusion in Cornish, Manx, and one of the Rumantsch varieties (Surmeiran?).

Incredibly off-topic here, but preocclusion in Cornish, Manx (and Norn!) isn't really the same as Surmeiran Verschärfung. The suface instantiations and underlying motivations are rather distinct. On the other hand, none of that contradicts the point that "weirdness of the fringe" is just as likely as archaism.

[8Deer]

I vaguely recall reading a paper (I think written or co-written by Johanna Nichols) that argued that different types of society had different grammatical features. It didn't say anything Wholfian, though, but claimed (and this is all from memory, it could be all different than I remember) that 1) languages spoken by relatively many people evolve quicker than languages that are spoken by only few people and therefore 2) smaller languages had more chance to retain some quirky features.

Other linguists have pointed out that it's not only easier to retain quirky features with a small and peripheral community of speakers but to innovate them as well. I remember years ago reading an article to this effect which took as its example pre-occlusion in Cornish, Manx, and one of the Rumantsch varieties (Surmeiran?).

Incredibly off-topic here, but preocclusion in Cornish, Manx (and Norn!) isn't really the same as Surmeiran Verschärfung. The suface instantiations and underlying motivations are rather distinct. On the other hand, none of that contradicts the point that "weirdness of the fringe" is just as likely as archaism.

Is this "Verschärfung" the same thing as "hardened diphthong" (http://en.wikipedia.org/wiki/Romansh_language#Phonology) mentioned on Wikipedia? I'd be interested in reading more.

[Drydic]

I vaguely recall reading a paper (I think written or co-written by Johanna Nichols) that argued that different types of society had different grammatical features. It didn't say anything Wholfian, though, but claimed (and this is all from memory, it could be all different than I remember) that 1) languages spoken by relatively many people evolve quicker than languages that are spoken by only few people and therefore 2) smaller languages had more chance to retain some quirky features.

Other linguists have pointed out that it's not only easier to retain quirky features with a small and peripheral community of speakers but to innovate them as well. I remember years ago reading an article to this effect which took as its example pre-occlusion in Cornish, Manx, and one of the Rumantsch varieties (Surmeiran?).

Incredibly off-topic here, but preocclusion in Cornish, Manx (and Norn!) isn't really the same as Surmeiran Verschärfung. The suface instantiations and underlying motivations are rather distinct. On the other hand, none of that contradicts the point that "weirdness of the fringe" is just as likely as archaism.

Is this "Verschärfung" the same thing as "hardened diphthong" (http://en.wikipedia.org/wiki/Romansh_language#Phonology) mentioned on Wikipedia? I'd be interested in reading more.

Verschärfung ~= sharpening

[Dewrad]

Is this "Verschärfung" the same thing as "hardened diphthong" (http://en.wikipedia.org/wiki/Romansh_language#Phonology) mentioned on Wikipedia? I'd be interested in reading more.

Yes, it is. Haiman and Benincà's The Rhaeto-Romance Languages has decent enough coverage.

[linguoboy]

I vaguely recall reading a paper (I think written or co-written by Johanna Nichols) that argued that different types of society had different grammatical features. It didn't say anything Wholfian, though, but claimed (and this is all from memory, it could be all different than I remember) that 1) languages spoken by relatively many people evolve quicker than languages that are spoken by only few people and therefore 2) smaller languages had more chance to retain some quirky features.

Other linguists have pointed out that it's not only easier to retain quirky features with a small and peripheral community of speakers but to innovate them as well. I remember years ago reading an article to this effect which took as its example pre-occlusion in Cornish, Manx, and one of the Rumantsch varieties (Surmeiran?).

And, if we define "quirky" as "not occurring in BIG languages"...

No need--unless preocclusion/Verschärfung is actually more common in "little" languages than we've previously suspected. How many other speech varieties can you name with this feature?

[Whimemsz]

I vaguely recall reading a paper (I think written or co-written by Johanna Nichols) that argued that different types of society had different grammatical features. It didn't say anything Wholfian, though, but claimed (and this is all from memory, it could be all different than I remember) that 1) languages spoken by relatively many people evolve quicker than languages that are spoken by only few people and therefore 2) smaller languages had more chance to retain some quirky features.

Other linguists have pointed out that it's not only easier to retain quirky features with a small and peripheral community of speakers but to innovate them as well. I remember years ago reading an article to this effect which took as its example pre-occlusion in Cornish, Manx, and one of the Rumantsch varieties (Surmeiran?).

And, if we define "quirky" as "not occurring in BIG languages"...

No need--unless preocclusion/Verschärfung is actually more common in "little" languages than we've previously suspected. How many other speech varieties can you name with this feature?

There's definitely some Austronesian languages with preocclusion, although I don't remember the names off the top of my head.

For verschärfung, don't forget Greek! There's also a very similar process of glide fortition (w > gw being especially common, but other types occur as well; found in a number of Austronesian languages, Blackfoot, Cheyenne, Piman languages, Armenian, Spanish etc. -- see also this post -- plus you might count the many instances of postlabial [j]/palatalized labials changing to labial+obstruent clusters), as well as the epenthesis of obstruents following high vowels, found in: Maru (Tibeto-Burman), Huishu (Tibeto-Burman), several Grassfields Bantu languages, Lom (Austronesian), and Singhi (Austronesian), and apparently some New Caledonian languages and Trengau Malay. (Source: "The Emergence of Obstruents After High Vowels," David R. Mortensen, Diachronica 29:4 (2012).) Mortensen also mentions that some Scandinavian dialects have this as well, as in some varieties of Jutland Danish bi "bee" = [biç]~[bik^j]~[bic].

[Radius Solis]

There's also the pre-stopping found in Icelandic, which what gave us [tl] medially and [tɬ] finally for <ll> in Eyjafjällajökull, which I hope I have spelled correctly.

[gach]

Pre-stopped nasals occur also in West Saamic, e.g.

N-Saa Sápmi < *šämä, cf. Fi Häme (endonyms)
N-Saa jiekŋa < *jäŋi, cf. Fi jää ("ice")

The stopping is written with homoorganic stops but is really more glottal.

[finlay]

There's also the pre-stopping found in Icelandic, which what gave us [tl] medially and [tɬ] finally for <ll> in Eyjafjällajökull, which I hope I have spelled correctly.

No ä in Icelandic, otherwise correct.

[brandrinn]

Most of the differences will be lexical, and some of them will be obvious based on culture and technology. Some of them might not be so obvious. Technologically primitive cultures in places like New Guinea often have very few words for color, for example, compared to the several basic color words of nearly all the major prestige languages of Indonesia, Australia, and Papua New Guinea.