Can animacy be grammatical or must it always be natural?

Question

Animacy has come up in a few recent questions, especially in comparison to gender. One interesting thing that turned up in the comments on those questions was whether or not animacy can sometimes be purely grammatical.

Let me compare with gender which is more familiar.

Natural gender and grammatical gender are two different things.

Natural gender is not abstract or arbitrarily assigned. This is when language reflects the sex of things, mostly people and higher animals. English has natural gender in pronouns etc: "he", "she", "him", "her", "his". Some languages have only one third person pronoun regardless of the sex of the person etc referred to.

Natural gender is not lexical. It refers directly to the sex of the thing referred to and not the word representing the thing.

Grammatical gender is abstract or arbitrarily assigned. This is when every word is assigned a class such as "masculine", "feminine", "neuter", "common". These assignments are mostly not related to sex but may be related to the sound of the word or may be utterly arbitrary. Words for inanimate objects are very often assigned to the "masculine" or "feminine" class and animate objects are sometimes assigned to the "neuter" (sexless) class, for example German "Fräulein" is a neuter word for "girl", not feminine as would be expected under natural gender.

Grammatical gender is lexical. It refers to the word representing the thing and not to the thing directly. Two synonyms referring to the same thing rrcan have different genders. For example there are two Spanish words for "star". "El astro" is masculine and "La estrella" is feminine.

It helps to differentiate the words "gender" and "sex" when talking about this topic. "Gender" was originally a grammatical term only and came to become a synonym for sex later.

There is another phenomenon often confused with grammatical gender. I might call it "semantic gender" but there's probably a real term for it. In English all ships tend to be referred to as "her", but this is not lexical, does not depend on the word used for the ship. Languages with grammatical gender may use different genders for "schooner", "destroyer", "carrier", and "battleship".

So, is animacy ever in any language grammatical? That is to say does it have the properties of being a) abstract/arbitrarily assigned and b) lexical?

Answer 1

In Russian, for example, animacy is not clear-cut in the distinction animate/inanimate. I'll explain better.

When studying Russian, or any language that have animacy, it's usually simple to distinguish animals and people that are animate, from a car or a table that are inanimate.

But there are some ambiguous cases.

Apart from animals and people, nouns that denote mythical characters and anthropomorphic creatures or mechanisms are usually animate (Group a), while nouns denoting plants, groups of people or animals (i.e. collective nouns) are inanimate (Group b) and lastly, other nouns can refer both to animate and inanimate nouns (Group c)¹.

A strange exception to this rule in Russian language are the words "мертвец" (dead person), "покойник" (the deceased) which are animate, while "труп" (dead body; corpse; cadaver) is inanimate.

An exception to the group B (collective nouns and plants) is пролетарий (proletariat) which is animate.

---

^{1: I'm not really sure about the Group C, because I found two slightly contradicting sources, unless I misunderstood that part. If some Russian native speaker can give a contribution, it would be nice.}

Answer 2

Algonquian languages are often used as examples of systems that use animacy as a grammatical category. For the most part, it falls into real-world animacy vs inanimacy, but there are cases where there are mismatches between real-world animacy and the grammatical category. This is always one-way - i.e., all real-world animate entities will be grammatically animate, but some real-world inanimate entities will be grammatically animate. (cf.Goddard 2001)

For example, in Blackfoot, the word for "knife" isttoan is grammatically animate, where real-world knives are inanimate. Another example of the arbitariness of the animate/inanimate distinction is the words for an earthenware vs metal bowl. Both types of dishes are ko's, but the earthenware bowl is grammatically inanimate, whereas the metal dish is grammatically animate. (cf. Frantz & Russell 1989/1995, Wiltschko 2009).

(I haven't answered unequivocally "yes" to the question, however, because some have argued that Algonquian nominal systems should not be divided according to animacy, but to a "noble"/"ignoble" or "esteem/no-esteem" distinction (cf. Maillard 1964 for Micmac). And Wiltschko, for example, has proposed that the animate/inanimate distinction in Blackfoot is morphosyntactically parallel to the mass/count distinction as opposed being morphosyntactically parallel to a gender distinctions.)

Answer 3

As in Russian, Polish marks animate masculine nouns by having a genitive-accusative vs. the nominative-accusative of inanimate masculines. There are some ambiguous cases, e.g. pilot 'pilot, TV remote'. A pilot is human and automatically animate. A TV remote is not, but it also gets the animate genitive accusative. Strangely enough, there are totally inanimate nouns like papieros 'cigarette, which take a genitive accusative: on pali papierosa 'he's smoking a cigarette'. We can write off the pilot case on temporal or cultural grounds, but the papieros case sounds like a treatment of animacy as a grammatical category. If you want to say this makes it no longer animate, I won't argue.

Answer 4

In Introduction to Classical Nahuatl, Andrews (2003) writes:

Since the Nahuatl nounstem is a labeling device, it implicitly carries a culturally established animacy classification of entity types. Because cultural factors are involved, a nounstem may have an animacy at variance with an English speaker's expectations; for example, Nahuatl nounstems translated by "hill" (or "mountain"), "heaven", and "star" are animate.

However, in modern Huasteca Nahuatl which I study, although cultural factors are changing rapidly, "hill" and "star" still take the animate plural form ("tepetl"->"tepemē", "citleli"->"citlelmē"). So what may have begun as natural animacy is becoming lexical animacy.

Answer 5

In Women, Fire, and Dangerous things: What categories reveal about the mind (1990), George Lakoff describes a tribe that places women, fire, and dangerous things into the same linguistic category. I read the book some time ago, but I believe that the tribe has a gender for animate objects. If so, then articles or tense or case would have to match the animate category, lest the usage be ungrammatical.

I think that what you call grammatical gender is simply called gender. They are really linguistic categories that are, as you say, somewhat arbitrary. Further, while many languages have two or three genders (masculine, feminine, neuter), there are languages that have many genders. In Chinese, for example, there is no masculine or feminine gender, so ta(1) can mean "he" or "she." However, there seems to be a separate gender for people san ge ren = three (people article) people. And another gender for money. san(2) quai(4) chen(3) = three (money article) dollars. These categories are very deep and not at all obvious to a non-native speaker such as myself. (Incidentally, I would welcome someone who knows or has studied Chinese to correct and extend my examples.)

Answer 6

I know that in older forms of Norwegian and other Scandinavia languages that there are to levels related to grammatical gender. There is animacy and then there is gender. Both of these are called gender but I think inappropriately. Within older forms of Norwegian there is the inanimate and the animate and then the animate is further divided into masculine and feminine. Most of these old forms have fallen into disuse and now remain only as inanimate and masculine. Most of the feminine form have become masculine although New Norwegian tries to maintain the feminine forms. I suspect that gendered languages either follow one or the other of these methods, or like Norwegian had both but have dropped one or the other. I am currently studying Abenaki, an Algonquian language which uses animacy but not true gender. Those things that are animate have agency and those that are inanimate do not. I think that understanding the two levels helps on understand this concept better rather than lumping it all under gender.