Nutshell 4. Some important sorts of words

Compound words

Rule One

A compound word is constructed from a two-word sentence – predicate and object of which become modifier and head of the compound, respectively – in the following way:

1. Prestem:
   1. the object’s prestem
   2. the object’s inner case
   3. the object’s poststem
   4. an optional separator: -, ~ or ^
   5. the predicate’s prestem
2. Inner case
3. Poststem: the predicate’s poststem
4. Outer case

Note that the object’s stem comes before the predicate’s; and also that the object’s outer case (and, less importantly, the predicate’s inner case) is lost. The separator can be used, for example, if the word boundary would be unclear otherwise, or for placing the second part of the word on a new line.

The object becomes the compound’s head, the predicate becomes the modifier, and the object’s inner case becomes the epenthetic case. We will symbolise compounding with ⇒.

Interlinear glosses of compounds consist of the head’s gloss, the epenthetic case, the modifier’s gloss, and the inner and optionally outer cases. In later units, in examples where compounds’ structures are not relevant, they are not always resolved.

This compound contains the information ‘the action of thinking is the X of having to do’. The missing case descriptor X has to be added from context. Thus, compounds are always a bit ambiguous.

Rule Two

In the relationship between the original predicate and object, the rules of sentence grammar are retained as far as applicable. (Forget this for the moment; we’ll get back to it in the next nutshell.)

Rule Three

Regarding all outward relations, cases (i.e. the compound’s inner case [not to be confused with its epenthetic case] as well as the outer cases of its objects) refer to the head.

A good way to think about this is that we first modify the head (we are not talking about a real action of eating but the permission to eat; we are not talking about a real action of running but the necessity to run), then add inner case and objects as if they were inner case and objects of the head (it’s about the location of eating, as opposed to the location of allowing; it’s about eating sweets, as opposed to allowing sweets; the running is fast, as opposed to the necessity; etc.).

Rule Three also means that the modifier cannot have any objects apart from the head. Put differently: if a main predicate has got more than one object, you can’t form a compound.

Examples

We can already express a number of subtleties with cumulative or partitive brackets or coordinations, yet compounding provides another possibility.

By inverting the bracket before compounding, the inner nominative of ‘waiter’ is preserved; if we did not do this, we would get bymÌwd. female-acc-waiter-acc¹. which could as well mean a female customer female-acc¹ waiter-dat-acc².. On the other hand, losing the inner accusative of ‘female’ is not a problem for comprehension. The inner case of the resulting compound, referring to the waiter, is of course also a nominative. This sort of construction downtones the modifier for reasons we will only get to on the pragmatics pages in the appendix. Other useful modifiers include numerals to form equivalents of singular and plural (see Grammatical number below).

The modifier can and should be omitted if the information it conveys is irrelevant, or if it is clear from context. Do not translate ‘waitress’ as mewdbè. waiter-nom-female-nom¹. or ‘brother’ as htryÌx. sibling-acc-male-acc¹. by default; the gender-neutral forms will do most of the time.

We will frequently mention that everything ‘irrelevant or clear from context’ has to be omitted. This rule is also explained on the pragmatics pages.

We need to make ‘cup’ the head so it can be sensibly used in sentences (‘I gave her a teacup’, where I gave her a cup, not tea) and take objects (‘a blue teacup’, where the cup is blue); but losing the inner accusative of ‘tea’ is not a problem.

Negators

The two negators are verbs conveying a property, or something rather similar to a property: nonexistence or opposition, respectively. They can thus be seen as adjectival verbs, with nonexistence or opposition in the accusative being conferred on the dative: nì. is something made into nothing, something undone; nÌ. is something with the property of nothingness, an action not done, a nonexistent thing. The difference between dative and accusative is more pronounced than with most adjectives: something that has never existed cannot be made nonexistent and therefore needs the accusative.

But then, the dative/accusative distinction is often lost again because we can easily form compounds with negators. The resulting sentences have lost the object’s outer case; they differ from their positive counterparts only in a suffix-like negation.

When we speak about an action being made undone, we actually mean its consequences: ‘unknot’ does not mean making the action of knotting nonexistent, but its effect, so we need the consecutive case.

The tortoise is an object of the negator, not of the knot-making: as a compound’s modifier cannot have objects, we can’t compound this sentence. If we don’t want to focus on the undoing of a previous action, we can use a compoundable phrasing with the opposition negator.

Negators are a good illustration of Rule Two of sentence grammar, namely the rule that an object is not simply a subordinate word, but a subordinate word plus all of its own objects. The above sentence, nà xOájy axileÌsi. not-fact¹ hear-fact-acc² Achilles-acc-dat^3a., does not just negate ‘listen’ but ‘Achilles listens’, leaving open the possibility of other people listening. nà xOájy axileysì lÌwOl. not-fact¹ hear-fact-acc² Achilles-acc-dat^3a lion-acc-psu³. means ‘It is not true that Achilles listens because of the lion [but he might still be listening for some other reason]’. Compounding the negator does not change this meaning, because Rule Two is of course still in operation.

Pronouns

Pronouns do not denote specific actions such as ‘to run’ or ‘to turn green’. They refer to actions, so to say, by pointing rather than naming.

Relative pronouns

The target of a relative pronoun’s stem is an action denoted by another nearby word stem. For example, w– refers to the stem of its own predicate.

The stem of this pronoun is a placeholder for the stem of its predicate wàx. ‘speak’, hence wì. here denotes ‘the recipient of speaking (the speaking named in the predicate)’. This is how we express English reflexive pronouns. This sentence is almost, but not exactly, the same as wáx wìxe. speak-fact¹ speak-dat-nom^2a.. The latter does not necessarily imply that the two ‘speakings’ are the same and thus means something like ‘The one being spoken to is speaking’. We will further examine the identity (or nonidentity) of actions in the next nutshell, in the chapter on Rule Four of sentence grammar.

Here is the complete list of the relative pronouns:

In other words: for a type I pronoun with some level n, start from that pronoun and move left until you reach the first entity (word or parole, which, as you recall, has level zero) with the specified level. For a type II pronoun, move left until you reach the second entity with that level.

In addition to reflexivity, these pronouns can be used to translate person, possessive adjectives (not to be confused with adjectives of possession), and other interesting things.

In interlinear glosses, ‘PI’ and ‘PII’ represent relative pronoun stems of types I and II, respectively, and the level is given in subscript. Hovering the mouse over the gloss also tells you to which action the pronoun refers – unfortunately, not on smartphones.

In the penultimate example, we want to refer to the sweet. We do this with a pronoun that points to the sweet’s predicate ‘eat’ and give it an inner case that matches the sweet’s outer case: an accusative. Therefore, the pronoun refers to the content of eating, which is the sweet. (Recall how we argued brackets.) This is an indirect reference or a reference via the predicate. The pronoun in the last sentence, by contrast, simply points to the tortoise from two words prior; this is a direct reference.

When a pronoun refers to a compound, compounding Rule Three (cases refer to the compound’s head) is also applicable to the pronoun’s inner case.

What has been stated of compounds’ modifiers is also true for pronouns: they can and should be omitted if the information they convey is irrelevant, or if it is clear from context.

Demonstrative pronouns

These two pronouns point to something in the world. We typically use them with an inner accusative just like nominal verbs.

Brackets with gwÌ. are usually partitive: gwÌ fkrÌjyn. any-acc¹ tortoise-acc-partacc². ‘anything from the set of tortoises’ is usually more appropriate than gwÌ fkrÌjy. any-acc¹ tortoise-acc-acc². ‘anything, a tortoise’.

Third-person pronouns are translated with definite or relative pronouns, in some cases with Ìx. ‘male’ or bÌ. ‘female’, or – and this is the best option – left out.

Numerals

Here are some important numeral verbs:

These are adjectival verbs meaning ‘make/become one individual, make/become two individuals’, etc. The inner accusative is again the most useful case. They form brackets or coordinations.

Weighting numerals with various outer cases translate adverbs such as ‘seldom’ — crÌaR. … 1/4-acc-temp². and ‘everywhere’ — jnÌar. … 1/1-acc-loc².. The nonexistence negator can be used likewise: ‘nowhere’ — nÌar. … not-acc-loc²..

Grammatical number

We can get close to the idea of grammatical number (singular and plural) by using a numeral as a compound modifier and thus downtoning it, analogously to the feminine in ‘waitress’ above.

Don’t be tempted to translate grammatical number like this on principle; in most situations, the number of things is irrelevant.

Nutshell 5