Unit 2. Words and sentences
The poem is never done.
The music never ends.
Not so long as you remember the words.
Not so long as echoes remain.
Word composition
All words are composed of the following parts:
Prestem + inner case + poststem + outer case
Prestem and poststem form the stem, or the lexical part, of the word. The division of the stem into two portions is similar to English verbs such as sing/sang/sung, where the lexical part is ‘s–ng’ while the vowels ‘i/a/u’ convey grammatical information. The stem always denotes an action (but never a state, a person, a thing, a property, etc.) and thus resembles our verbs. We will use the term ‘verb’ interchangeably with ‘word’.
Lemizh stems correspond more or less to the type of verbs called eventive (‘laugh, spit, speak, happen’), not to stative ones (‘sit, owe, be’).
The prestem can contain any sounds, or it can be zero (i.e. consisting of zero sounds). The poststem can only contain fricatives and plosives, or it can be zero as well.
The inner case is represented by one of the eight vowels, optionally followed by a liquid (the so-called primary case suffix) and/or a nasal (the secondary case suffix). The outer case has the same structure. For the first word in each sentence, the main predicate, the outer case is zero.
Level of words
There is only one more grammatical category: the level of a word. A word can be of first level (the highest), of second level (the next highest), of third level (still one level lower) and so on; there is no limit for the number of levels, but non-positive word levels (zero, −1, etc.) are forbidden.
The first word in a sentence (the main predicate, as you remember) is of first level by definition. The level of the next word is determined by the main predicate’s accent and by the type of pause between the two words, the level of the third word is determined by the accent of the second and the pause between these two, and so on.
Here is the complete list of pause/accent combinations. (The meaning of an agentive level will be explained in unit 3; it is not important right now.)
Following pause | Accented vowel | Type of accent | The level of the next word is … |
---|---|---|---|
space | inner case | low | lower by 1 |
high | lower by 1, and agentive (abbreviated a ) | ||
outer case | low | equal | |
high | higher by 1 | ||
comma (,) | inner case | low | higher by 2 |
high | higher by 3 | ||
outer case | low | higher by 4 | |
high | higher by 5 | ||
full stop (.) | inner case | low | none; end of sentence |
There are no unstressed words (clitics) in Lemizh.
In interlinear glosses, words are given in their English rendering, followed by the abbreviations of inner and outer cases (which are listed below) and a superscript number for the level, with an optional a for ‘agent’. Pronoun stems have special abbreviations starting with a ‘P’; they are treated in unit 6.
Now look at these examples and follow the word levels from the main predicate to the last word: (Remember you can hover the mouse over the text to display its transcription.) | |
---|---|
wáx kRenseè laÌ hemÌsi. | |
speak-fact1 Trance-nom-nom2a do-fact-acc2 Seamus-acc-dat2. | |
là xt xOà jy hnà xty zèi, hemÌse jhèjy. | |
want-fact1 hear-fact-acc2 fate-fact-acc3 PIn−3-nom-dat4 Seamus-acc-nom2 machine-nom-acc3. |
Rule One of sentence grammar. Sentence structure
The word levels determine the structure of a sentence.
All words of second level are subordinate to the main predicate (which has first level). A word of third level is subordinate to the next second-level word in front of it, and so on. The main predicate itself is subordinate to the parole, the action of speaking (or writing) the sentence in question, which consequently has level zero. Think of it like this: instead of saying ‘X’, you can say the equivalent ‘I say X’ — then ‘I say’ is the parole.
Here is a schematic ‘sentence’ with the words represented by their level numbers: | |
---|---|
The strict formulation of Rule One is as follows: A word of level n is subordinate to the nearest word of level n−1 in front of it; the parole acts as a word of level zero.
Rule Two. Definition of objects
An object of a word in a sentence is a word that is subordinate to the former, plus all of its own objects. In the diagram above, the main predicate’s three objects are enclosed in ellipses. The first object of the main predicate is not only the second word, but consists of the four consecutive words with levels 2, 3, 3, and 4. This may not seem important at the moment, but we will get back to it later. Objects of the same word are called sibling objects or just siblings, and the word they are subordinate to is their predicate. The first third-level word in the example sentence and the following third-level word (plus its fourth-level object) are siblings with the preceding second-level word as their predicate. Note that ‘predicate’ and ‘object’ are relative terms like ‘parent’ and ‘child’.
Objects name certain circumstances of their predicate’s action: its sender, recipient, location, cause, and so on.
The parole acts as the predicate to the sentence as a whole; and the sentence is the parole’s object. This, too, will only become important later.
And here is the strict formulation: An object of a word in a sentence is a word subordinate to the former, its predicate, plus all of its own objects.
Rule Three. Outer case
What exactly does an object say about its predicate? Or, to use a technical term: what is the thematic relation of an object to its predicate? This is determined by the outer case of its first word.
Outer case works very much like case in Indo-European languages such as Latin or German. Unfortunately, most English words do not distinguish case, so many native speakers are not familiar with the concept. English pronouns, however, do make this distinction.
Therefore, pronouns serve nicely to illustrate the point: | ||
---|---|---|
wáx veè bÌi. | I speak to her. / I tell her. | |
speak-fact1 PIn−2-nom-nom2a female-acc-dat2. | ||
wáx byè vèi. | She speaks to me. / She tells me. | |
speak-fact1 female-acc-nom2a PIn−2-nom-dat2. |
‘I’ and ‘she’ are the ones who do the speaking or telling; they are subjective or nominative forms. In Lemizh, the case which denotes the sender of the information is also called the nominative; it is represented by the vowel e. ‘Her’ and ‘me’, the recipients of the information, are (in Lemizh) dative cases, marked by an i. The accusative, as we will see in the following example, denotes the content; it is represented by an y. In general, each case is defined by its so-called descriptor (sender, recipient, etc.), which characterises the object’s relation to its predicate (or, strictly speaking, to its predicate’s stem, which denotes an action).
We will sometimes show Lemizh cases in the translations like so: | |||
---|---|---|---|
wáx byè nagcÌ vèi. | She is the sender, the war is the content, and I am the recipient of the telling. | She-nom speaks about a war-acc to me-dat. She-nom tells me-dat about a war-acc. | |
speak-fact1 female-acc-nom2a war-fact-acc2 PIn−2-nom-dat2. |
The strict formulation is: The outer case of the first word of an object defines its relation to its predicate’s stem via its descriptor; the outer case of a level 1 word (the main predicate) is zero.
The remaining four rules will have to wait until later.
Case descriptors
Here is a table showing all primary cases, those without a secondary suffix. Don’t try to memorise all of them right now; we will get to most cases during this tutorial. The secondary cases, those with a secondary suffix, are treated later.
No. | Case vowel | Primary case suffix | |||
---|---|---|---|---|---|
none | l | R | r | ||
Plot cases | Causal cases | Temporal cases | Spatial cases | ||
1 | a | factive (fact): action | affirmative (aff): fact (point in causal chain) | temporal (temp): time | locative (loc): place/region |
2 | e | nominative (nom): source, sender | causative (caus): direct cause | ingressive (ing): starting time | elative (ela): starting point/region |
3 | y | accusative (acc): content | contextual (ctx): causal context | durative (dur): duration | extensive (ext): spatial extent |
4 | i | dative (dat): sink, recipient | consecutive (cons): direct consequence, effect | egressive (egr): closing time | illative (ill): end point / ending region |
5 | o | tentive (ten): intention | intentive (int): intention (intended point in causal chain) | episodic (eps): episode, ‘act’ | scenic (sce): scene, ‘stage’ |
6 | O | comitative (com): company | persuasive (psu): reason | digressive (dig): time away from which | ablative (abl): place/region away from which |
7 | u | instrumental (ins): means, tool | motivational (mot): motivational context | progressive (prog): time that is passed | prolative (prol): crossing point/region |
8 | U | benefactive (ben): beneficiary | final (fin): purpose, aim | aggressive (agg): time towards which, temporal aim | allative (all): place/region towards which, spatial aim |
A predicate has exactly one nominative (sender), one dative (recipient), and so on. Many cases are not stated explicitly (not overt) most of the time: none of the above example sentences names its persuasive (reason), but nevertheless implicitly has got one – in a very broad sense: it can also be ‘for no reason at all’, but it will still exist. On the other hand, there cannot be two senders, recipients and so on. There is, however, nothing hindering the recipient from being a set of several people (‘Dylan and Seamus’). This is our first glimpse of the later rules of sentence grammar. We will keep coming back to this point.
We will use ‘nominative’ more or less synonymously with ‘nominative object’. The first term focuses more on the relation, the second more on the grammatical structure.
Inner case
Just as the outer case describes a word’s relation to its predicate, the inner case describes its relation to its own stem. In the sentence wáx veè bÌi. speak-fact1 PIn−2-nom-nom2a female-acc-dat2., the nominative object (‘I’) describes the sender of its predicate ‘speak’. Correspondingly, wèx. speak-nom1. – the stem of the word ‘speak’ plus an inner nominative – is the sender of speaking, the one who tells something, the speaker. Just as with outer case, wèx. is not necessarily a single person; but I will not mention that every time we encounter a noun.
wìx. speak-dat1. is the recipient of speaking, the one who is told something. wÌx. is the content, the tale. wà rx. is the place where something is spoken or told. wà x. is the action of speaking, equivalent to our infinitive ‘to speak/tell’ as in ‘You are not allowed to tell lies’, or the gerund ‘(the) speaking/telling’ as in ‘Telling lies is bad behaviour’. Infinitives and gerunds will become important towards the end of this tutorial. The inner factive of some verbs can also be interpreted as an abstract noun with gerund-like meaning: ‘the warring = the war’ (see the example above), ‘the helping = the help, the behaving = the behaviour, the controlling = the control’, as well as emotions such as ‘the loving = the love, the fearing / the being afraid = the fear, the being happy = the happiness’. (Being happy or afraid isn’t simply a state like being tall or brown-haired. Consequently, these verbs are expressed by word stems in Lemizh.)
In unit 4 we will learn how inner case is used to form concrete nouns (‘tail’) and adjectives (‘purple’).
Objects of ‘nouns’
You already know that the outer accusative of a word denotes the content of that word’s predicate.
But what about this noun phrase? | |
---|---|
wèx nà gcy. | |
speak-nom1 war-fact-acc2. |
Of course, the war is not the speaker’s content, but the content of the speaking: the phrase means ‘the one speaking about a war, the teller of a war’. Have a look at the exact wording of Rule Three: the outer case defines the relation to the predicate’s stem, which is an action.
Inversion
There is no particular word order between sibling objects: they can change places whenever desirable. It is a different matter with words in a predicate-object-relationship, since the predicate comes first by definition. However, the main predicate and one of its objects can change places by a process called inversion (symbol: ⇔), swapping inner and outer cases of the initial object.
là cw wèxi. | [Someone] helps the speaker. | |
help-fact1 speak-nom-dat2. | ||
⇔ wáx lìcwe. | The one being helped speaks [something]. | |
speak-fact1 help-dat-nom2a. |
There is a symmetry behind an inversion: in our example, both sentences claim that the sender of speaking is the recipient of helping. The equation is wèx. = lìcw., the speaker = the one being helped. This might seem like playing around at the moment, but we will encounter inversion in most of the units to come.
The above sentences also illustrate that you can omit any object in a Lemizh sentence (such as the nominative ‘someone’ or the accusative ‘something’) without the risk of getting an ungrammatical sentence. You might get an ambiguous or even incomprehensible sentence, though, if you omit too many objects.
In other words: there are no complements (necessary objects), only adjuncts (optional objects).
Any objects of the two inverted words have to be taken along.
là cw wèxi nà gcy3. | Someone helps the teller of a war. | ||
help-fact1 speak-nom-dat2 war-fact-acc3. | |||
⇔ wáx licwè nà gcy2. | (‘war’ remains an object of ‘speak’) | The one being helped tells someone about a war. | |
speak-fact1 help-dat-nom2a war-fact-acc2. |
Inversion normally works only between the main predicate and one of its objects. Trying to invert a lower-level word and one of its objects would leave us with a word having two predicates, which is technically impossible. The one exception will be treated in unit 5.
From the viewpoint of a native Lemizh, inner case is inflection, not derivation: wà x. ‘to speak’ and wèx. ‘speaker’ are forms of the same word, like ‘speak’ and ‘spoke’ in English.
Partitive cases
From each of the primary cases we can form a corresponding partitive case (abbreviated part) by adding the secondary case suffix n: thus, en represents the partitive nominative, arn the partitive locative, etc.
Partitive cases are, of course, defined by their descriptors, which are a bit abstract: ‘the set from which the sender (recipient, place, etc.) is thought to be taken’ or more informally ‘the type of the sender (recipient, place, etc.)’. These cases correspond more or less to English ‘some (of the), one of the’.
xacgà zw meqxèn hemÌsi. | The stars are the set from which the sender of twinkling is thought to be taken. That which twinkles is from the set of stars / is of the type ‘star’. | Some of the stars twinkle at Seamus. | |
twinkle-fact1 star-nom-partnom2 Seamus-acc-dat2. | |||
Strictly speaking: since the partitive only states that the twinklers are from the set of stars, but not how many of them, we cannot exclude the possibility that all of the stars twinkle at Seamus. The point is that the above sentence does not demand that all stars twinkle at Seamus. That’s another issue we will keep coming back to.
The other secondary case suffix, m denoting the qualitative cases, will be treated in unit 11.
Inner partitive
Now we can use an inversion to find the meaning of the inner partitive cases.
là cw wèxin. | The speakers are the set from which the recipient of helping is thought to be taken. | [He] is helping some of the speakers. | |
help-fact1 speak-nom-partdat2. | |||
⇔ wáx lìncwe. | They who are being helped are from the set / of the type ‘speaker’. | The one(s) being helped, among others, speak. | |
speak-fact1 help-partdat-nom2a. |
Hence the inner partitive translates as ‘among others’. If this doesn’t make sense, draw an image like the one with the stars above.
Again, we cannot rule out the possibility that all of the speakers are being helped. The strict translation is ‘(possibly) among others’, but the point of the inner partitive is that it opens the possibility that some of the speakers are not being helped.