Taekrand

Phonology
I use the Romanised letters in the following tables. Although an alphabet does exist, it is syllabic (and as yet undecided).

Consonants
The letters C and K are (unfortunately, due to my own inconsistency) used interchangeably. Note that G and C are always 'hard' (ie. [g] and [k]; never [s] or [dʒ] etc.) - although in the case of C it is easier to use K than risk mispronounciation by English speakers.

H is very rare, only occuring in (to date) one word. It is a non-word-initial allophone of HY or HU.

The voiced-voiceless distinction is non-contrastive - some words and/or dialects prefer voiceless stops, others voiced, though voiceless is much rarer. If the stressed syllable ends in a voiced stop, a nasal is added between them (as in taekraNd) - this is written in the romanised letters but is not considered phonemic by native speakers. Note that the two approximants HY and HU are voiceless.

Phonotactics
The fluidity of the language is a priority for me, so clusters of obstruents have to be restricted to allow the liquids and vowels to take priority. Basically, consonant clusters of two or more fricatives, and three or more obstruents (of any sort), are forbidden and and consonants are lost from the root until this is achieved. Liquids are retained, as long as they are not between the two obstruents. Thus the genitive of 'taekrand' is 'taekdē' (ie. taek[r][n]dē), and the adjective from 'shachar' (dryness) is 'ashar' (ie. ash[ch]ar) (dry).

Basic Grammar
The six vowels, arranged in order, give rise to various grammatical distinctions like case, aspect etc. It is important to know the order, as most of the time the 'endings' are not formed by the use of a specific vowel, but rather 'shifting' vowels along the sequence. Thus, if for example a verb has its last vowel one 'shift' lower than usual, it can be deduced that it is in the future tense. The vowels, in order, are:

It is possible to mark vowel shifts using accents. A shift up (ie. towards the 'i' end of the sequence) is marked by an upwards-pointing 'up' accent (ie. a circumflex), while a shift down is marked by a 'down' accent. Sometimes a grammatical form always uses the same vowel; this is called a 'flat' shift and is marked by a 'flat' accent.

(Unfortunately due to computer limitations 'w' cannot be accented, so a bit of guesswork is required. It helps to know that a flat shift to 'w' does not exist outside the relative case. Also for 'ae' the accent goes on the initial 'a')

So:


 * 'e' shifted up goes to 'î'
 * 'a' shifted down goes to 'ǒ'
 * a flat shift to 'e' is written 'ē'

Note that the letter 'w' is relatively unstable and sometimes forms use 'o' where 'w' would be expected and 'w' where 'o' would be expected.

Other forms that are important to recognise are inverted and compound forms. Inversion means that the position of a vowel and adjacent consonant are switched (ie. C-V → V-C or vice versa. Here C is a consonant and V a vowel.) Compounds have two vowels; typically this looks like a word which has been inverted and then had the original (or other) vowel added back (so V-C → V-C-V, for example). In both cases shifts can and often do affects the vowel(s) as well.

Eastern Dialect
In the eastern dialect, spoken most of all by the Laes Hyaerra, the following pronunciation changes are found:


 * 'Hy' and 'hu' become 'y' and 'w' respectively
 * 'S' becomes 'z'
 * 'Th' becomes 'dh'(voiced)
 * Glottal stops are made into 'h's Yozhik: Strictly speaking ' isn't so much a glottal stop as a diaeresis or similar, so that mu'at is pronounced [muat] (as opposed to [mwat] or something) and so'ot is [sɒːt] (as opposed to [sut] or [sʊt] or something)
 * Alveolar plosives become dental plosives

Yozhik: Actually the two are not contrastive in western dialects and which is which, even in English, still confuses me. So this may be a nothing change, or else I can force (specifically) alveolar in western dialects.

Napitok: At least the way I'm saying it, there is an audible difference, and one that would certainly make you sound slightly different in English (if you swapped all the alveolar 't's and 'd's for dental ones). But I'll show you what I mean when we next meet.

Yozhik: It's not just as straightforward as dental vs alveolar vs postalveolar like we all thought, there's actually laminal vs apical vs subapical, palatalised vs velarised, dental vs interdental, fricatives vs sibilants, etc. I might try and do a Pronunciation page with tables and stuff on to explain it all...

Eg. so'ot in the Orliaean dialects would be pronounced zohot in the eastern version, and 'Laes Hyaerra' would be 'Laez Yaerra'.

Grammar
Taekrand is a highly inflected language, where most prepositions and particles used in English are conveyed by some sort of case, aspect, affix or other form. This makes it very concise and so sentences are often considerably shorter than in other languages, or indeed English.

Roots
All words in Taekrand are based on a root of (usually) three consonants. In between these consonants are (usually) two vowels. Depending on the position and shift of these vowels, different parts of speech can be derived from the root. For example, the word 'pucor' - land - has the root 'p-c-r', from which are derived the following:


 * 'upocr' (adj.) - of the land, natural
 * 'pucor' (adj.) - real, 'worldly'
 * 'pucor' (noun) - land, country
 * and others.

Almost all roots are triliteral - that is, they have three letters. (This is simply because there aren't enough two-letter roots, but four-letter ones are too long to be convenient for everyday words.) Some roots only have two letters, for example those of the words 'mis' ('m-s') = 'thing'; 'sot' ('s-t') = 'in existence', "is". Other words appear to belong to this group, but actually a consonant has 'dropped out' because the word was too difficult to pronounce. For example 'ashar' = 'dry' has the root 'sh-ch-r', as is evident in words like 'shachran' = 'a dry thing', "the dry". However, although the consonants are enough to identify the root meaning, the vowels are required to identify its grammar when shifts are concerned. To use the same example, the vowels in 'ashar' are 'a-a', so the essential noun is 'sh-a-ch-a-r', the adjective is 'a-sh-a-(ch)-r', and so on. A shift down in the word would give the vowel 'o', so 'oshar' is the ergative case; this could not be deduced if the root vowel were not known, as it would be impossible to determine the 'direction' in which the shift had taken place.

Nouns
Yozhik: Revised to be more intuitive and easier to understand. Please re-read if you found the previous version confusing.

Nouns in Taekrand have eight cases: these are nominative, ergative, accusative, ablative, allative, genitive, relative and locative. Of these, the last three are 'oblique' and are only used to add extra information, while the first five can be used for arguments of the verb. There are four numbers: singular and plural, as in English, quantitative expressing a part or amount of something (eg. 'pishtâe' (qu.) = 'some water') and collective expressing a group of individuals that form a unit. For example 'lapesǔ' (cl.) = 'trees' would more naturally be translated as 'forest' (ie. a group of trees that functions as a unit).

* An alternative form here is 'pucwr'. This is used more often in speech, but is identical to the ergative singular

As a summary of an otherwise complicated table:


 * The singular has no specific mark. No compounds are used
 * The plural uses a vowel with shift down. All the cases except nom., erg. and acc. use compounds.
 * The quantitative uses a vowel with no shift. All the cases except nom., erg. and acc. use compounds.
 * The collective uses the flat shifted vowel 'e'. All the cases use compounds.

And:


 * The nominative has no vowel shift in the singular and quantitative, and a shift down in the plural and collective.
 * The ergative always shifts down, and the accusative always shifts up
 * The ablative always uses 'a', the allative always 'i' and the relative always 'w'
 * The locative never shifts, but can take inversions/compounds
 * The genitive uses 'e' or 'o' depending on the number.

Strong and Weak Nouns
This is a distinction of form, not meaning. Although it cannot be predicted, the following are often weak rather than strong:


 * Abstract nouns
 * Long or technical words
 * Words in a vowel high up in the sequence
 * Some women's names*

* This use of so-called 'weak' words is NOT a reflection of the speakers' views on gender!

Weak nouns swap the endings of pairs of cases: erg. swaps with acc.; abl. with all.; and (singular only) gen. with an archaic form (which has disappeared from modern Taekrand) called the Absolute, which ends in x↓i - that is, like the weak accusative with an 'i' on the end forming a diphthong. So, for a weak noun like 'kutak' = 'hate', in the singular:


 * the ergative is 'kutâek'
 * the accusative is 'kutǒk'
 * the allative is 'kutkā'
 * the genitive is 'kutkǒī'

...and so on. Care needs to be taken not to confuse these shifts with their corresponding strong forms. Also note that in the plural, in a strong noun the nom. is the same as the erg., but in a weak noun it is the same as the acc.

Because of the pairs of cases that swap, it is likely that many weak nouns are weak because they are more often used in the accusative than the ergative. Thus the shift down, which usually results in a richer sound that may well have been thought of as more emphatic than an up shift, is used more often. (It is from this idea, that the ergative form of these nouns was less emphatic than the equivalent form in other words, that they got the English name 'weak'. The huinglu and laes hyaerra who speak the language, however, apparently have no word for it, and happily borrowed the English one when they needed it!)

Actually, that isn't quite true: the reason some words are weak and others aren't depends on the Shisad Language's pattern of 'complementary vowels' - that is, vowels were arranged in pairs and rather than shifting up (or down) they swapped to the other vowel in the pair. When the pairs were linked together to become a sequence in Proto-Taekrand, a noun was 'strong' or 'weak' according to which direction in the sequence its complementary vowel lay. As the idea of direction became more important than that of complement (ie. when the ergative case joined the accusative in being represented by a shift - suddenly vowels had two possibilities to change to) more and more nouns changed the direction of shift in these cases to conform to other nouns of similar meaning. Thus the nouns that have stayed weak and defied 'regularisation' to a strong form tend to be those that are either so common that any such change is too difficult to slowly integrate into speech, those which are so rare that they aren't used enough to have a chance to be changed or would simply be misunderstood if they were, or those for which there would be more ambiguous endings (where an up/down shift ending clashes with a flat shift one) if the direction were to change. Equally some, though a smaller number, have gone the other way.

Alignment
This is confusing, with lots of subtleties, and the general pattern must be quickly learnt to make sense of the language. Apart from the oblique (gen., rel., loc.) ones, all the cases in Taekrand can be used as the argument of a verb. For now I will ignore the two 'indirect' cases (all., abl.) as they complicate the picture. This leaves nom., erg. and acc.

The purpose of the noun cases is primarily to convey emphasis and verb voicing distinctions before the noun's grammatical role in the sentence. As such, just about any case can be used for any argument. However, the distinction between the subject and object is made by the following rule:


 * The case of the subject (agent) is always 'below' that of the object (patient).

By 'below', I mean 'in a case that lends itself more strongly to the role of object (patient) in a sentence'. Namely, accusative is below nominative; nominative is below ergative. This system of case hierarchy is quite important.

Beyond that, the use of cases is largely free for emphasis and voicing purposes. Now I will talk about each alignment system in turn:


 * Basic Alignment: Tripartite This means that the nominative case is used for the argument of intransitive verbs, accusative for the object of transitive verbs, and ergative for the subject of transitive verbs. (cf. Na'vi language; if the na'vi wikipedia page has a better explanation, let me know.) This alignment carries no voicing emphasis, and applies when both subject and object are given equal weight.


 * Secondary Alignment: Austronesian So called because it's found in Austronesian languages, apparently. This is an extention of the above system that allows a particular argument to be weighted more than the other. Here the nominative case is used for the argument with the most weight. So nom-acc has active 'antipassive' weight, while erg-nom has 'passive' weight and corresponds to the passive in English. Note that the hierarchy rule is still obeyed.


 * Nom-Nom Alignment This is occasionally used to give pronounced weight to both arguments, usually to emphasise (confusingly, as they are unmarked) that the subject and object are not reversed (as, perhaps, may have been the addressee's misconception). Here the word order (SOV) and/or the differential(s) distinguish subject from object. (Since no other emphasis is needed, word order can be used for this)

Pronouns
These are used very differently to in English. There is no first, second and third person, nor does the pronoun used change depending on the speaker and the addressee. Instead, replies will usually use the same pronoun to refer to the same people as was used previously. The two most important prepositions for use in conversation are 'dos', which is reflective, and 'caeth', which refers to the narrator. This is made clearer by examples:


 * If you are talking about yourself or telling a story which you are in, refer to yourself as 'caeth', since you are the 'narrator'.
 * If you are asking a question, and the reply will be as described above, refer to the addressee as 'caeth'. When they reply, they will use 'caeth' of themselves, which doesn't involve anyone being referred to by a new pronoun.
 * If you are asking a question about yourself, use 'dos', even though you are the narrator, as using 'caeth' would prompt the addressee to reply using 'caeth', which to them would refer to themself. Their reply will also use 'dos', but this does not refer back to them (despite it being reflexive) as you have already used it for yourself.
 * If in doubt, the (indeclinable) pronoun 'thi' can be used in conjunction with any other pronoun, and is uniquely second person. However it is usually used for formal and polite questions, so it can sound odd if used with someone you are close to.

'caeth' and 'dos are both regular in declension

Narrative Pronouns
These are used for referring to people when telling a story. The two most important pronouns are 'i' and 'oha'*. As with all other pronouns, the person they refer to should not be changed. Their uses are outlined in more detail below:

* This is the only occurrence of the letter 'h' in the language (apart from 'hy' and 'hu'). The laes hyaerra dialects tend to pronounce the h as in English, ie. [h]. But the huinglu dialects tend not to pronounce it at all - although they don't introduce a glottal stop [ʔ] or make it into a diphthong either.


 * The 'hierarchy' is that 'i' is 'higher' than 'oha', so it refers to the most 'important' character: it is called the primary pronoun. 'Oha, the (you guessed it) secondary pronoun, refers to the second most 'important' character. While tertiary and subsequent pronouns do exist, they are considerably rarer, and identical to the equivalent ordinal (eg. shosad, taersad).
 * In a long narrative, 'i' always refers to the same person. 'Oha' usually does too, though its use is more flexible. (Tertiary and subsequent pronouns are generally avoided and almost always used as a one-off, whose character is understood from the context, and is likely to be different from the previous use and to change again before the next one.)
 * Alternatively, and this is more common in speech where many people are being referred to simultaneously, they can be used like (but much more commonly than) 'the former' and 'the latter' in English; thus 'i' is the last person mentioned, 'oha' is the person last mentioned before 'i' and so on. (Although it is more common to name a person again than use tertiary and especially quaternary pronouns).

'Oha' only has one number, singular, although it can refer to plural or collective nouns. It is as regular (strong) as any word ending in a vowel can be. Where inversions would normally happen, stick an 's' on the end - so the locative is 'ohas', etc. The ablative ends in 'o' (ie. 'ohos') so as not to clash with the locative. 'I' is highly irregular, though loosely weak. The singular and quantitative have a stem in s, although this is not seen in the nominative, as well as several unexpected diphthongs. The plural has a stem in r, while the collective borrows that of 'mis' = 'thing', which despite its stem is regular (strong) in itself: 'cērǎ' / 'cǎr' (from irregular plural 'crǎ').

Verbs
Verbs in Taekrand fall into three main categories. The first is 'Verbs Of Action', which are most similar to English's verbs and have the familiar systems of tense, aspect, and so on. 'Verbs Of State' covers all those verbs which describe a state of being, such as 'to live', 'to know', etc. All the grammar for these verbs is exactly the same as for adjectives. 'Verbs Of Change Of State' are a form of Verbs of State which describes the action that caused that state to come about. They are dealt with in their own section.

Verbs in Taekrand belong to one of two tenses (present and past), three moods (indicative, weak and strong conditional - that's not including infinitive) and three aspects (imperfective [habitual], perfective, and aspect future - that's not including continuous, which is formed from the infinitive, or the perfect participle). Mostly, any combination of these is possible. The basic tense/aspect classification system is done by shift, so that the present forms are a shift up, the future forms a shift down, and the past forms unchanged. (This means you should always look up the past infinitive in a dictionary!) The mood is formed by inversion, so that the indicative is formed by inversion so that it ends in a vowel (eg. 'naecrw', from root 'naecor' [=flight], 'muta' [=do, verb-only root]), and the infinitive is formed by inverting the othe way, so that the two vowels are adjacent (eg. 'nae'wcr', 'mu'at'). In addition, there is a past suffix 'li'. The exact meanings vary between moods, so each is looked at separately:


 * In the indicative (almost) all aspects are possible in all tenses. The present tense is unmarked; the past technically adds the suffix 'li', although it is equally acceptable to omit it. The imperfective is used more or less as in English (x does/did [regularly] / used to do), as is the perfective (x did [on one occasion] - this only exists in the past tense). The future aspect is used only with reference to the current tense (: x is/was going to do) - it refers to a time after the one which the main story/speech refers to.
 * The imperfective is formed by a shift up from the root (eg. 'naecrô', 'mutâe')
 * The perfective has no shift (eg. 'naecrw', 'muta'). It is only used in the past tense (although the suffix 'li' is, as ever, optional)
 * The aspect future has a shift down (eg. 'naecrǔ', 'mutǒ').
 * Any of these can be made past by adding the suffix 'li', or by context.
 * The infinitive is used as in English (ie. 'to do x'). Between this and the participle, which has identical form, they can fill the role of any noun and adjective as well as verb, declining using the differential particle. Here the shifts mark tense, not aspect, so that the present infinitive has a shift up, the future one has a shift down, and the past one is umnarked.
 * The participle, identical to the infinitive, is also used to get round the lack of an imperfective (continuous) aspect by saying literally 'I am/was/will be a doer of x' ie. 'I am/was/will be doing x' by using the same shifts as the infinitive as described above, (optionally) agreeing it with the subject by way of the differential.

Taekrand has no passive (or any other) voice. Instead, different combinations of cases are used: erg-acc for normal emphasis; nom-acc for active/antipassive; erg-nom for passive (see also nouns.) Also the word order can be changed to bring the focus to the front of the sentence for emphasis, or delay it for tension and dramatic effect.

Adjectives
Adjectives fall into two main declensions: thematic and athematic. This is a difference of form, not meaning, although patterns of meaning can be observed among adjective pairs derived from the same root but in different declensions. Thematic adjectives have a thematic vowel (ie an 'x') and decline by doing shifts and inversions to that vowel, while athematic ones simply 'borrow' the vowel and inversion from the noun they qualify. In both athematic and thematic adjectives, all such shifts and inversions happen to the initial vowel; thus 'asha(ch)r' (=dry) becomes 'aeshar' in the accusative, 'shichar' in the allative, etc; and 'pucor' (=real) uses forms like xpucor, px'ucor etc. (There are a small number of adjectives, most notably those numbers which decline, which shift the final vowel, like nouns.) Almost all adjectives Adjectives have only one number, so they only agree in case. In athematic adjectives, however, for numbers outside the singular, inversions take place as if the noun was singular. Thus 'ashar odbǔ' = dry people; ǔpucor (not pǔ'ucor) odbǔ = real people). The vast majority of thematic adjectives are strong; however there is a small group that is weak, and an even smaller group that can be either strong or weak, agreeing with the noun in 'strength'. (In athematic adjectives the distinction is meaningless, as the vowels do not in themselves determine the case, only which noun they qualify.)

Because the verb 'to be' does not exist (in a transitive sense, at least) an adjective agreeing with a noun is enough to convey the idea of being [something is something]; you can get a past tense [something was something] by adding '-li' to the end, as for verbs.

A lot of adjectives in Taekrand correspond to English verbs, such as 'to exist' = 'sot' (adjectival meaning: 'in existence'); 'to know' (aware of); to sleep (asleep), etc. Their grammar is exactly like any other adjectives.

It is also possible to 'back-form' a perfect participle from any verb (the opposite to the process of forming change-of-state verbs from adjectives). This has the (adjectival) meaning 'in a state resulting from having done x / been x-ed' (or some neater translation where one exists, or simply 'having done x / been x-ed' when one doesn't). These can be active or passive depending on the context; all For example:


 * 'Dos nacr' - 'I am in a state resulting from having flown' or simply  'I have arrived (by flying)' / 'I am here'
 * 'Âet mut' - 'It is in a state resulting from having been done' or simply 'It is done / finished' / 'It has been done'

To form these, the two vowels must be contracted to form one vowel. The new word behaves like a monosyllabic root (ie. CVC). The rule for contraction is:


 *  'The new vowel is one shift away from the first vowel in the direction of the second vowel' 

Thus: nae'wcr > na↓cr, ne'isk > ni↑sk, shw'ek > sho↑k

Note that if the resulting vowel is w, it often shifts back to the one it came from, thus mu'at > mu↑↓t, not mw↑t.

Numbers
Taekrand uses a base-8 number system. The numbers one and two decline like nouns (one is weak; two is strong and of course plural), and three like an adjective. Numbers greater than four can be declined using the differential where necessary. Ordinals are formed by adding '-ad' to the end (this becomes '-sad' after a vowel (ie. shisad, shasad) and '-d' after kwna and dwna (ie. kwnad, dwnad). At any rate they are easily recognisable).

For reference the numbers are as follows (the larger ones are in standard form):

NB. The pattern for the bigger numbers is: The number x×8n would be written no-x; so 6×87 is seveno-six (which is: luno-dws)

Subcognatives
The following affixes can give extra detail about the role, strength of meaning, or specific-ness of a noun, verb or adjective:

The first six are positive; the last three are negative. The three negatives correspond to the last three positives (th ≠ eish / so ≠ ma / is ≠ col). Note that with subcognatives 'mis' (=thing) uses the (strong) stem soc, giving eish(s)oc, colsoc etc. (and the irregular definite msol).

The negatives are all translated with 'not' in English in different ways, for example:


 * thdakw i aet-cos = he didn't say that it was finished
 * sodakw i aet-cos = he said that it wasn't finished / he denied that it was finished
 * thdakw i měs = he didn't say the thing
 * sodakw i měs -or- dakw i sosâc = he didn't say anything
 * thso'ot = not to make
 * thsot = not in existence / not having come to be made
 * isso'ot = to destroy / to 'un-make'
 * issot = no longer in existence / having come to be destroyed

Note that the three positive subcognatives can all be mae using double negatives:


 * thmuta i th(s)âc = he did not-nothing = he did something
 * thmuta i sosâc = he did anti-nothing = he did anything
 * thmuta i issâc = he did all but nothing = he did everything

Change-Of-State Verbs
Yozhik: I did these last 3 sections in a bit of a hurry so if you think they aren't detailed enough please say so (write so?) and I will add more detail

These are verbs that describe the action that caused a state to happen. They are formed by doubling the vowel (eg. 'sot' > 'so'ot', 'cos' > 'co'os')

They can be transitive or intransitive depending on the combination of cases used. Some are transitive-orientated - that is, a single argument in the nominative is the subject, with the object unspecified. This means that by default the sentence structure is nom-acc; the reflexive pronoun 'dos' can be used as the accusative object. Others are intransitive-oriented - that is, a single argument in the nominative is both the subject and the reflexive object. Here the default sentence structure is a monovalent nom; it can be made transitive by adding an agent in the ergative. There is no pattern to which orientation a COS verb has.

Differential
This particle has the following uses


 * To clarify ambiguous case-endings where the context isn't sufficient
 * To emphasise the grammatical role of the noun it agrees with over its meaning
 * To indicate the grammatical role of a noun when an unexpected case ending is used (eg. 'âet-taecor dakw = the word was said; cf. 'taecâr dakw' = something said the word)
 * To supply a noun where its grammatical role is important but not its meaning (eg. 'cos i âet' = he/she finished it; cf. 'cos i' = he/she are finished

It is hyphenated to the word it agrees with. It declines as follows:

Derivation
A taekrand root can form lots of words. The most common and universal forms are:


 * Active = this is the verbal/adjectival form, and is formed by simple inversion. The meaning is basically the same as the root meaning. The adjective form is always thematic.
 * Essential = this is the noun form, and is also formed by inversion (or rather, the lack of it). The meaning is the state described by the adjective (corresponding to the suffix '-ness') or the process described by the verb (corresponding to the suffix '-ation') which is the active form.

And, when the active form is adjectival:


 * Athematic Adjectival = the meaning of this is often more abstract, but related (eg. 'upocr' = 'of the land'; 'earth-y' but 'pucor' = 'real' [cf. 'worldly']). Formed as expected.

Any other meanings are formed by creating a so-called secondary root by adding an extra letter. There is little pattern here, but some common letters are:


 * Consequential (aka subjective) = the typical subject of the verb, or noun described by the adjective; the 'doer'. Formed by adding perhaps ~n, ~l (eg. Taecor [=word] > taecrol [speaker]; Ash(ch)ar [=dry] > shachran [=a dry thing])
 * Objective = the typical object of the verb or adjective. Forms vary
 * Active = the process/means of doing the verb, usually a structure, system or process. Formed by adding perhaps ~d, ~↑d (eg Taecor [=word] > taekra↑nd [=language])