Detail #422: Variations on Reflexives

I have been reading up a lot on reflexives in different languages, and this inspired me to write a little on the different types of reflexives I have encountered, and some additional types.

1. Types of pronouns 

1.1. Gaps

Some languages have gaps in the pronominal system where the reflexives would be expected to appear. West Germanic, for instance, now entirely lacks reflexive possessives (whereas Slavic and North Germanic do have them).

Thus, "he saw his car" can signify the subject's car being seen by the subject, or some other third person's car being seen by the subject. In Slavic and North Germanic, there is a reflexive possessive pronoun. In examples throughout this post, I may use "sy" and "sine" for this, analogous to this structure w.r.t. Swedish: min:my::sin:sy.

Another arguable gap in most IE languages is the lack of a reflexive subject, which could make sense with subclauses and such:

"he did not know that heself would ..."

There are other imaginable situations where a reflexive nominative also makes sense, such as in split ergative languages where the nominative also is the absolutive, or in languages where some type of quirky case or differential case marking sometimes has nominative forms in object positions. Also, in the bit further down about reference, we may find other reasons.

Another imaginable gap could be a gender-specific conflation. Imagine a language where masculines and feminines have a distinct reflexive form, but neuters do not, leaving

"the animal saw it"

 

ambiguous as to whether it's reflexive or not.

2. Person

In some languages, reflexive pronouns are person-specific, whereas in some they are entirely person-agnostic. Russian is an example of the latter, English of the former.

Thus, in Russian, reflexive arguments often are expressed by the pronoun 'sebya' in the proper case form regardless of person (although in first and second person, using the first or second person pronouns is permitted and sometimes done). In English, it's myself, yourself, himself, herself, etc.

In Swedish, first and second person use the first and second person pronouns. The semi-reflexive "själv" (obvious cognate) can follow, but is optional.

jag såg mig in the mirror = I saw myself in the mirror
jag såg mig själv i spegeln = I saw myself (emphatic) in the mirror

I would go so far as to say that "själv" no longer is properly reflexive in Swedish but rather some kind of intensifier and restrictive marker. C.f.

Jag visste att jag själv skulle bli tvungen att lösa det.
I knew that I self would have to resolve it

Han själv hade inte hunnit med det, men med hennes hjälp gick det bra.
He self had not been_able_to_do_on_time with it, but with her help went it well
He would not have been able to get it done on time by himself, but with her help it went well (or maybe "alone, he would not ...")

Otherwise, 'själv' serves the other roles 'self' serves in English, altho' sometimes in the superlative: självaste kungen/kungen själv : the King himself.

The usual reflexive in Swedish only pops up in the third person, and does not distinguish number - although 'själv' would be inflected for plural if used with a plural, and neuter when used with neuters:

han såg sig (själv) i spegeln
de såg sig (själva) i spegeln
djuret såg sig (självt) i spegeln

he/they/the animal saw him/them/itself in the mirror, -a = plural adjective/pronoun congruence, -t = neuter ditto.

The richer the congruence system on the reflexives get in a language, the more likely it feels like the reflexive/non-reflexive distinction is going to be lost and be replaced by some form of proximative/obviative-like distinction instead. Once you have gender and number and the whole shebang both on the regular third person pronouns and reflexive ones, you will very seldom need a restriction on them that force them to be reflexive or not, but rather might just care about whether there's two referents that are distinct and of somewhat different prominence in the discourse.

3. Reference

3.1 Subject only 

It is not unusual for reflexive pronouns to be restricted to subjects only. Thus, sebya and its forms, as far as I can tell, only refer back to the subject. I have no idea how this works with non-nominative subjects of infinitives in Russian, but there's space for variation there.

Natural variants of this type could be absolutive-only and topic-only.

3.2 Some other kind of reference

In Swedish, the rules for the possessive reflexive are complicated, but as an acquaintance of mine would express it: 'Any NP that C-commands the phrase with the reflexive pronoun can be the possessor'.

In fact, there's a further sort of restriction where for most speakers, the regular pronouns cannot refer to the subject (whereas the reflexive ones can; the non-reflexives do seem to be able to refer to non-subjects even in positions where they can refer to non-subjects as well.) 

Two examples:

de visade honom till sitt nya rum

they showed him to sy new room

Here, the new room might be theirs or his.

John visade Peter till hans nya rum

John showed Peter to his new room

Here, the new room can only be Peter's.

The c-command rule opens up a few other positions:

tanken i all sin förträfflighet
the thought in all (of) sy excellence = the thought in all its excellence

This could be any constituent - subject, object, prepositional object, etc. This position, however, does seem to permit for use of non-reflexive pronouns as well,

tanken i all dess förtäfflighet
the thought in all (of) its excellence

Finally, the Swedish reflexive can refer to an empty subject of an infinitive, and this also holds for the regular reflexive object pronoun

att lära sig känna sina gränser är viktigt
to learn to know sy boundaries is important

att vila sig är hälsosamt
to rest oneself is healthy

Beyond this, there is a nominalization in the plural, 'de sina', which signifies the closest family and friends.

3.3. Even further kinds of reference 

The Swedish example above is rather complicated, but we can find examples that it does not permit. Deciding whether to permit these (or to restrict some of the ones the Swedish example includes) can give some space for a conlang to grow into a detailed project.

3.3.1 Conjunctions 

"Han och sina vänner" - "He and sy friends" does not work in Swedish. It must be "han och hans vänner".

The reflexives do not work over subclause boundaries, making

"han visste att han var försenad"
he knew that he was late

mildly ambiguous as to whether it is reflexive or not. "Själv" could be added, but would sound really weird in Swedish - though more acceptable with some other verbs in the main clause and subclause. Even then, "själv" is not necessarily reflexive, as it might actually also signify that the second subject is or does something by himself.

3.3.2 Topicalization of non-subjects 

One could also imagine that topics are possible candidates for reference of reflexives, in which case you might want to be able to refer to the subject as possessed by the topic, and with a reflexive possessive at that. Or maybe even in some weird situation where the topic and the subject are the same referent, but mark different roles,

himself he gave an expensive gift.

 

could make much more sense in some languages' logic as

 

him heself gave an expensive gift.

 

3.3.3 Quirky case 

 

Of course, with quirky case you may have a non-subject in the nominative or a subject in a non-nominative case, and there may be restrictions depending on whether they're true quirky cases or not affecting whether they can or can't be the referent of a reflexive. It might be nice having objects  in the nominative be candidates for reflexive reference (and also, naturally, blocking the use of regular reflexive possessive pronouns owning the object).

3.3.4 Subordinate structures

 

It is imaginable that subordinate infinitives with an agent that is distinct from the subject of a finite verb may restrict the reflexives within its scope to refer to NPs within the infinite VP + the agent, or maybe even more restrictive, such that the agent may be blocked from being the referent of the reflexive. Thus

I helped him do his homework

could, in such a language, not be

I helped him do sy homework

 

One could also imagine a reverse effect, where the agent is within the scope of the infinitive phrase's block and prevents external reflexives from reaching it, thus if he helps his sister with her homework, it couldn't be

he helps sy sister do ...

But if the infinitive then permits reflexive reference to the agent, this would be permissible:

he helps his sister do sy homework

The situation with subclauses is of course of some interest as well, but I will not get into detail with regards to that. Similar possibilities exist as with regards to the infinitive phrase, but with a subclause you generally do not have arguments "outside of" the scope of the subclause.

4 Other considerations

 

4.1. Distinct reflexives for subjects and other referents

 

One could imagine a language that has evolved distinct forms of 'sy' and '...self' for subjects and objects (or ergatives and absolutives, or topics and non-topics, or subjects and objects and obliques, or topics and other NPs or maybe topics or subjects vs. non-subject non-topics). This seems unlikely but not impossible. I imagine this would likely include some kind of morphological marking distinguishing the two (or more) types instead of separate stems.

 

4.2 Antireflexive 

 

 I have previously considered an anti-reflexive pronoun, where the regular third person pronoun is assumed to be reflexive if there is a third person subject, and any non-reflexive reading requires an explicitly non-reflexive one.

4.3 The Finnish reflexive possessive

 

The Finnish reflexive possessive is fascinating in that it entirely lacks independent morphemes of its own. Normally, reflexiveness in Finnish is marked by itse + case + possessive suffix, so "itselleni" = self + to + my = to myself.

However, 'itse' does not feature in reflexive possession. 'Oma' can be used for that sometimes, but is not exclusively reflexive. In meaning it is fairly close to '(one's) own', i.e. more about exclusive possession rather than reflexive such, although both meanings do exist.

Third person reflexive possession in Finnish is expressed by the possessive suffix, whenever the noun is not the subject:

Hän löysi varastetun autonsa = (s)he found her/his stolen car

 

Here, we actually get a slightly antireflexive construction, because if you sneak in the genitive third person pronoun, it suddenly no longer is reflexive

Hän löysi hänen varastetun autonsa = (s)he found her/his(someone else's) stolen car

Ŋʒädär: Indefinite Address

Indefinite address in Ŋʒädär differs from that of English significantly. Indefinite address does reuse parts of the indefinite pronoun system for some constructions - but only because the dedicated indefinite second person address pronouns lack certain case forms.

1. Indefinite 2nd person

Besides the usual second person pronoun vär, Ŋʒädär has a rather special second person indefinite pronoun, 'jusa(n)' (absolutive), 'jusam' (dative). It has a rather simplified case morphology, but has a specialized morphological system. It seems fairly clear it originates with the imperative "jus", listen up.

1.1 Use

The pronoun is used when addressing (at least) one individual out of a group, such that the speaker is not aware of the identity, but is able to deduce the existence of, or at the very least suspects the existence of, a person that fulfills some given criteria. In writings, it an also address any reader that has some quality, or any reader in general. With the spread of literacy, it has especially taken to being the term of address employed when instructing any reader to do something - in letters to a specific reader, the second person is used instead.

1.2 Morphology 

The pronoun only distinguishes two cases, the absolutive and the dative. Other cases are conflated either with the second person pronoun vär, or some  indefinite pronoun (depending on context, style, time, personal preference of the speaker, etc).

However, jusa(n) has some special morphology, with some amount of syncretism in the system. It is similar to the indefinite pronouns lisar and nusar, with the exception that lisar and nusar have a full case system (with some syncretism).

	adnominal	adattributal	adclausal
absolutive	jusar, lisar	jusada, lisada	sajusan, salisan
dative	jusam, lisam	jusada, lisada	jusam, lisam

Forms such as jusaŋa, jusus, jusuk, jusluno etc do appear in speech, but rather infrequently. They do seem to elicit a certain sense of "wrongness" whenever used, both in most hearers and speakers. 

The adnominal can refer to an adjective or a noun.

jusar ŋator ('someone fast (among you)')
jusar kamma ('a/the chieftain (among you)')

lisar ŋator (someone fast)
nusar ŋator (something fast)

If it is known that at most one such individual can exist, the 2nd person plural possessive often marks the noun or adjective, i.e.

jusar kamma-un ('your chieftain' - assumed to be present)
jusar ŋator-un ('the fastest person among you')

Sometimes, the complement case is used both for adjectives and nouns:

jusar ŋatoɣ-ɣuv (in northwestern: jusaɣ ŋat:wuo)
jusar kammo-ɣuv (in northwestern: jusaš kaŋ:wuo)

In early modern Ŋʒädär and still in northern and northwestern Ŋʒädär, this marks a weakened certainty of the presence of such a person. In central and southwestern, it has rather come to be used with irrealis verb forms and questions.

Challenge: Person-like Things

Person is quite common in conlangs, but we can even almost find other person-like structures in human languages - some Asian pronoun systems rather may be ranked by social stature than by speaker > listener > other.

We could imagine some other system, though - the simplest being some way of adjusting how social stature is determined (age, with gender as tie-breaker? gender, with age as tie-breaker? some other type of social structure than any known human structure); maybe to some not entirely 'ordered' type of ranking (so it's more of a social network than a social hierarchy).

Suggestion: come up with a different system than person, and post in a comment.

Detail #291: Personal Pronouns with Complications

Few conlangs complicate up the third person pronoun a lot. Certainly such conlangs exist, but I still think it's reasonable to make up a short sketch of a language where personal pronouns are significantly more complicated, in order to inspire more such conlanging.

Let's consider a language where grammatical gender does not exist. There's one primary distinction, however – either animate-inanimate or human-nonhuman. I'll go with the latter. So, we now have pronouns for humans (let's use the Finnish pronoun 'hän' as a stand-in for this), and non-humans (let's borrow from Finnish again, 'se'). 

However, the language permits using other pronouns for third persons along a variety of distinctions. There's no necessity for there to be a pair of contrasting pronouns, even. We could imagine a pronoun for 'masculine' existing without a pronoun for 'feminine'; thus 'hän' could refer to either a man or a woman, but 'he' would refer specifically to a male. We could imagine certain age groups having pronouns, people of certain statuses or maybe in certain relative ranks of status, etc. Different pronouns could make different number distinctions as well. 

What could we do to make such words sufficiently unlike nouns to justifiably deem them pronouns? A few strategies exist:

• they can mark different grammatical categories
• more or fewer case distinctions (esp. convincing if nouns almost fully lack them)
• more or less specific number distinctions
• more suppletion
• lack of definiteness marking
• differences in distribution of articles and the like (i.e. they do not take counters or articles at all)
•  syntactical differences
• require antecedent or other referent to bind to
• follow the same binding rules as other pronouns with regards to what nouns they can refer to
• may not permit having adjectives or even attributes at all; may just restrict what kinds of attributes they accept (e.g. no relative clauses)
• may be cliticized as 'almost-congruence' markers on verbs in ways that nouns don't; sometimes, appear doubly due to this.
• tend to keep referring to the same referent after the first use, even if other possible referents surface - other third person pronouns that may fit the bill for the new referent will be used
• form reflexive pronouns
• have class-specific indefinite pronouns for each of the kinds of indefinite pronouns that the language distinguishes for indefinite pronouns in general, whereas nouns behave in a less complicated way with regards to indefiniteness of that type

Detail #31: Pronouns doubling as particles

Certain pronouns - especially indefinite ones - lend themselves easily to adverbial uses as well, c.f. how some languages do not distinguish "not" from "no, none". (Example: Swedish dialects in southwest Finland use 'inga' for basically all negation, thus having the same negation in "I didn't see you" and "No car is fast enough". )

Somewhat inconsistently, dialects of Swedish also use 'all' to denote completion (and by extension, wearing out, lacking and running out: bɪɭɪn jä all - the car is worn out, mjöɭt͡ʃen jä al: - IMD at least usually used to say something like 'well, that was the last of the milk!'. The pronoun has gender marking in congruence with the noun. This is basically only used with all as a complement - a rather restrictive use. How about constructing a whole perfect/perfective/telic(?) aspect using all as the morpheme/particle marking it? I'd presume this use would not have congruence with anything (or whatever form the pronoun takes would be influenced by whatever.)

Certainly, of course, something along the lines of some could easily be an explicit marker for slightly intensifying a imperfect/imperfective verb. Do keep in mind that different languages of course may have different sets of indefinite and definite pronouns (outside of the basic personal pronouns), there's no guarantee that 'some' will have a perfect match.

'who, which, that, ..' (relativizers): questions like "Who built this?" could easily get an increasingly grammaticalized answer like "it's John who [built it]", where "who [built it]" slowly is reduced to just an intransitive relativizer. This would end up increasingly like some kind of definite article restricted to only some contexts, such as answering questions, contrastive statements ("... but it was really he who."), etc. (Reminds me of lambdas in programming, for some reason).

Just some ideas that could be developed, and probably are rather close to what actually may have happened in some natural languages.

Coming up at some point in the next weeks: a consideration of pragmatics in conlangs.

Detail #26: Small set of vocabulary items

In a language with about English levels of inflection (well, maybe German, rather), a set of particles used for short subclauses along the lines of

... as did I

... than you are

... than he/she/it was 

 ... and so did you

... and so, he was [...]ed ... 

... and so, I ...

... than me  or ... than I, both distinct as they would be if the prescriptivist idea that the complement of than is subject or object of an omitted VP were correct. 

These would be formed from pronouns with verbal endings on them, with maybe some morphemes that are not used in any regular way in other verbs, coding for transitivity (do vs. be), type of use (comparison vs. inclusion ("as did I!")  vs. temporal sequence (and so, I ..., followed by inflected verb) vs. maybe some other usages), voice (active vs. oblique? active vs. passive vs. oblique? Where oblique stands for the pronoun being anything but subject or object;

As for voice, it would be nice if that were marked unlike how voice normally is marked in the language, and distinguished the two or three voices somewhat differently from the other voices in the language. I also imagine gaps in the paradigm would exist - first, the voice and transitivity markers would clearly be somewhat fusional, and I figure intransitive would mark obliques with the same morpheme the transitive equivalents use for direct objects. However, I also find the idea of having, say, some person lack some specific combination of use * transitivity * voice to be very appealing, like, maybe, lack of transitive oblique temporal sequence for first person or somesuch.

Historically, I imagine this would come about by reduced verb forms assimilating worn down pronouns in short subclauses.

Detail #18: Weird things to do with adverbs

In some languages where adjectives are verb-like, let adjective-verbs form auxiliaries that mark all kinds of manner. Even permit numbers to form auxiliaries that mark "to do n times", or "to do the n:th time" or "to do to N objects" (in effect extracting the number from the object NP to the VP).

When many adverbs are called for, some may remain chained with the main verb, and if several are highly salient they can be coordinated with a shared complement VP along these lines:

John quicks and carelesses stack[infinitive] the books

John carelesses stack the books and quick

John carelesses stack the books quick

John stacks the books, carelesses and quicks

If the adverbial has little to do with the subject, they may take a dummy subject:

John's house burned down, [dummy] unfortunates. 

Maybe a causative would be nice there?

[dummy] unfortunate.causative.3sg John's house burned down.

Would maybe work best in a language where infinitives can mark subject agreement? Alternatively, a language where there is some but rather sparse verb morphology, e.g. something along the system in standard Swedish with an explicitly marked present tense distinct from the infinitive.

As for numbers, some pronouns could also work:

I many.1sg have ideas - I have many ideas, "I profusely have ideas"

You howmany.2sg run laps? - You ran how many laps?

They how.3pl.past? - How did they do it?

 On the other hand, if done badly - without making a really thought-through morphosyntax here, it would easily end up a silly way of marking things in the wrong place.

Detail #5: Intrinsic, implicit referents

Does any natural language use obviative pronouns also to refer to nouns that have not been mentioned as such, but which are kind of intrinsically related to prominent nouns in a sentence?

Something like, say,

children look like them.obv = children look like their parents
house had no him.obv = the house had no master
sheep.plur know.3pl him.obv = the sheep know their shepherd

Not all nouns, necessarily, have their own associated nouns such as these; some have many, differentiated by gender and numbers.

When this is supposedly "normally" obviative or an indirect reference like this may be unclear from context.

Also, challenge #1 still stands. I have personally made no progress on it.