Today I realised one thing: my blog is subtitled "Arts and Crafts of Words and Tongues", but I've got only two posts tagged "language", and absolutely no post about conlanging, despite it being my second most favourite hobby! (everybody will agree that my most favourite hobby is procrastinating). So I've decided to fill this hole by writing a series of articles about my conlangs, starting with a few grammars (which will be spanned over a few articles). Doing so I'm actually killing two birds with one stone: not only I generate a bit of contents for a blog that desperately needs it, but I also create more exposure to my invented languages, for which until now information has only been available on the archives of the Conlang Mailing List, or in French.
To inaugurate this series, I've decided to publish the grammar of the Moten language, starting with a background introduction and its phonology (and related fields, like phonotactics, morphophonemics and the writing system). Moten is one of my first languages, and to this day still the most developed one (it has even got a lexicon!), so it's the best place to start. So without further ado, here is the first part of the Moten grammar!
Background Information (External)
Moten is one of the "First Five", i.e. the last five languages I had invented before joining the Conlang Mailing List, and the first ones that were not shamefully bad and forgettable. Of them, it is the most developed one, and still my favourite. According to my notes, I originally created this language because I wanted to create something that looked as if it had a lot of irregularities, but was actually very regular (the seemingly irregular surface forms actually being the result of completely regular morphophonemic mechanisms). Also, I had just learned some Basque grammar, and wanted to emulate some of its features (mainly the periphrastic conjugations and the fact that only the last word of noun phrases gets declined). Finally, I had discovered infixes, and wanted to see whether I was able to develop a conlang using them. Whether I managed to reach those goals is arguable. In any case, the result was a language not quite like anything I had created so far, and something I still think has got lots of character.
Background Information (Internal)
Languages don't always need a fictional background. Some are invented "just 'cause". Actually, that's how Moten started. But in time, a kind of weird fictional history crept in, without me trying to consciously create it. It just appeared in bits and pieces as I was designing the language, and eventually became too strong for me to ignore. It's this internal history I'm presenting here:
On the 25th of March 1984, a boy was found wandering in the Walloon countryside by a couple of Belgian farmers. He was naked, looked hungry, and despite being visibly frightened, he was so exhausted he didn't resist when the farmers took him and brought him to the nearest hospital. There, he was found to be about eight years old, and, despite the ordeal he seemed to have gone through, to be relatively healthy. Still, he was kept in observation as he had the symptoms of a profound shock: he refused to speak, was afraid of everyone and everything, and suffered from panic attacks and nightmares.
His picture was broadcasted, but without success. After two months, nobody had yet contacted the police. During this time, the boy slowly recovered, until he started to speak again. That's when the second mystery began: the boy didn't speak French, and nobody seemed to recognise his language. Interpreters were brought in, but all in vain. People started wondering whether he was a feral child. However, a speech therapist who was working in the hospital and had taken a liking for the boy quickly recognised that he didn't suffer the usual impairments of feral children: he walked upright, didn't display any sign of animal behaviour, ate normally, and his language ability sounded well developed, although his language was still unknown. Also, the boy showed signs of picking up French words without even being encouraged to, another clue that his language capacity wasn't limited.
With the help of a child psychiatrist, the speech therapist started testing the boy's intelligence using non-verbal tests. The results were surprising: the boy scored very high on those tests, showing a superior intelligence! Also, he began to learn French at an accelerated rate, although at first he had difficulties with some of the sounds. The speech therapist helped him, and he started getting attached to her. After a year and a half, he left the hospital, and the speech therapist got custody of him. She ended up fully adopting him a year later. In this time, he had learned to speak French with only a slight foreign accent, and went to school like all children his age (he had to learn writing and reading first, but got up to speed very fast). Despite him being able to communicate, his past remained a mystery: he had no recollection whatsoever of his life prior to being found, and that amnesia resisted all treatments. But that didn't matter, as he was starting to get a normal life.
However, he was still known as "the child that had been found in the fields", and his celebrity was in the way of his social development. To try and provide him with a normal childhood, his adopted mother decided to leave Belgium. She changed their names, found a job in France, and settled there in anonymity. As it would happen, she settled in a town close to where I lived, and one day, I met the boy.
Although I had no idea about his past, our encounter was still a shock: we were nearly exactly alike, as alike as identical twins could be! That was not all: our birthdays were identical! (in his case, the day he was found was used as his birthday) Those coincidences created a near-instant bond, and we became the best of friends. In time, I learned all about him (at least what he remembered, which was still nothing prior to him being found in the Belgian fields), and started getting interested in the language he spoke prior to learning French (yes, even in this fictional history, I'm still a language geek!). That's how I discovered that not only he still had a perfect recollection of his language, but also a very good idea of how to describe it grammatically, far better than what you would expect for a guy who presumably lost any contact with other speakers of that language when he was eight.
It took me a while, but I eventually convinced him to write down as much as he could about his language, arguing that the only link he had to his mysterious past was something too precious to let it disappear. C.G. (that's his nickname, what this means exactly will be revealed in a future post) was reticent at first, but eventually relented, and the result was the first description of the Moten tongue this world has ever seen. I also started to learn the language, and together we started speaking it, creating new compounds and loaning words from other languages for concepts (especially place names) that didn't seem to have already existing equivalents in Moten.
General Language Information
Now that this is out of the way, let's focus on the Moten language itself. Moten is, as far as anyone knows, a language isolate with a very non-Indo-European grammar (but a very simple phonology for generic European speakers). It is strictly head final (especially verb final), although attributes normally follow the noun. It is generally SOV and pro-drop, although not to the same extent as Japanese. Apart from two auxiliaries, verbal conjugation in Moten is strictly periphrastic. Word formation is mostly done by compounding, although Moten has a few derivational suffixes. Nouns inflect for case, number and definition. Fully inflected nominal and verbal forms can receive further nominal inflections to change their meaning or their role in the clause (a phenomenon usually called by its French name: surdéclinaison). Inflection uses a combination of suffixes and infixes, and sometimes also prefixes. When those combine with nouns, various phonological phenomena can happen, but those are always regular. Moten is a very regular language, with irregularities few and far between.
Moten has a relatively simple phonemic inventory with little to no allophony. It has 5 vowels and 18 consonants.
Moten's vowel inventory is simply the 5 main cardinal vowels: /a/, /e/, /i/, /o/ and /u/, like the Spanish and Modern Greek languages. As in those languages, the two mid vowels /e/ and /o/ in Moten are truly mid, i.e. they are slightly lower than the ideal versions described by their IPA characters. They would be more correctly transcribed as /e̞/ and /o̞/ (or as /ɛ̝/ and /ɔ̝/), but the versions without diacritics are used for simplicity. Moten also lacks diphthongs (two consecutive vowels always have a syllable break between them) as well as long vowels (except in a few interjections, but it is common for those to break the phonology of a language). Vowels are only very slightly nasalised when followed by a nasal consonant, and nasalisation isn't phonemic anyway.
Moten's consonant inventory is about as boring as its vowel inventory. It is relatively symmetrical, and features sounds common in most Western European languages. The following table shows the full inventory (using the IPA):
|plosive||p b||t d||k g|
|fricative||f v||s z|
When two phonemes appear in the same cell, the first one is voiceless and the second one voiced. Phonemes appearing alone in a cell are considered voice-neutral, but are normally realised as voiced. Moten's consonant inventory lacks any rhotic consonant (loanwords with rhotic consonants normally replace them with /l/). Also, while most coronal consonants are alveolar, /t/ and /d/ are realised as dental. As I wrote above, phonemes in Moten usually have a single perceivable realisation. The only exceptions are /n/ and /l/, which are usually realised as [ŋ] and [ɫ] when followed by a velar consonant (but this is a rare occurrence). /l/ is otherwise always clear in any position (unlike in English where it darkens in codas).
Due to its peculiar history, it is unknown whether the Moten language has a native script (if it has one, eight-year-old C.G. hadn't learnt it yet, or forgot it as part of his amnesia). So when C.G. and I started talking about writing a grammar of the language, the question of a written transcription came early. Eventually, C.G. created an alphabet for it, based on the Roman alphabet. The simple phonemic inventory allowed him to create a phonemic transcription where one letter is exactly one phoneme (the system is truly phonemic: morphophonemic sound changes are always reflected in writing) without having to divert much from the standard Roman alphabet as used in French or English.
The Moten alphabet has 23 letters, corresponding exactly to the 23 phonemes of its inventory. Of those, 19 are taken straight from the Roman alphabet, while 4 were invented specially for Moten. However, even those can be approximated in typography by digraphs using the pipe | followed by another letter. This is the convention used here.
The following table lists all the letters of the Moten alphabet in alphabetic order, in capital and small form, followed by the phoneme they transcribe in IPA, the name of the letter (used when spelling), and an example Moten word starting with the letter.
|A, a||/a/||a||at: fire|
|B, b||/b/||ba||bazlo: town|
|D, d||/d/||da||di|la: mother|
|E, e||/e/||e||elej: sleep|
|F, f||/f/||efa||fuli: gold|
|G, g||/g/||ga||ge|sem: father|
|I, i||/i/||i||ibo: air|
|J, j||/j/||eja||jem: brook, river|
|K, k||/k/||ka||ku|lu: language|
|L, l||/l/||ela||linan: bird|
||L, |l||/ʎ/||e|lo|||la: peace|
|M, m||/m/||ema||mosu: paw|
|N, n||/n/||ena||nudel: respect|
||N, |n||/ɲ/||e|no|||nuba: letter, character|
|O, o||/o/||o||oskana|not: ceremony|
|P, p||/p/||pa||pe|la: to see, to watch|
|S, s||/s/||esa||sigoj: name|
||S, |s||/ts/||e|so|||suko: brother|
|T, t||/t/||ta||tina: room|
|U, u||/u/||u||umpi: house|
|V, v||/v/||eva||vone: (cold) water|
|Z, z||/z/||eza||zanej: (finger) ring|
||Z, |z||/dz/||e|zo|||zika: mountain|
As you can see, Moten lacks the letters c, h, q, r, w, x and y. Those are not used in loanwords either. Rather, loanwords are adapted to fit the phonology (and in some measure the phonotactics) of Moten.
Moten has a relatively strict syllable structure (not as strict as for instance Japanese, but stricter than English). Syllables, no matter their position in the word, have the following shape: (C)(C)V(C), i.e. an optional onset of one or two consonants, a single vowel, and an optional coda with no more than one consonant. In particular, sequences of two vowels are always separated by a syllable break. Agreeing with this syllable structure, up to three consonants can appear between vowels. Syllabification is then done with a preference for the onset of the second syllable, although clusters of two consonants are separated equally between the syllables.
Vowel sequences are quite limited in Moten, despite the fact that a syllable break is always present between two vowels. Basically, only a, e and o can freely follow each other. I and u are never allowed next to another vowel.
Consonant clusters have other limitations besides number of consonants allowed. The main rule is that consonants in clusters always agree in voicing (the nasals, laterals and approximant are considered voice-neutral for this purpose, and thus can appear next to voiceless consonants as well as voiced ones). The second rule is that two identical consonants never appear next to each other (even across a syllable boundary). The third rule limits the occurrence of the consonants |l, |n, |s and |z. Those behave sometimes as a single consonant, sometimes as if they were clusters of two consonants. So they can appear in syllable codas, except when they are followed by a consonant, onset of the following syllable. When they appear in an onset, no other consonant is allowed in that onset. Note that this rule is not always followed in loanwords. Since they were all adopted at a time when C.G. could speak other languages (like French and English) with freer phonotatctics, some loanwords allow those consonants to appear in situations where they are disallowed in native words. An example is the word Doj|slan: Germany, which was loaned from German Deutschland. It is syllabified as Doj.|slan, which shows a |sl onset that is normally disallowed in Moten. If the word had been borrowed following all the phonotactic rules, it would have been *Doj|sulan, but C.G. and I just don't use this form. Notice however how the ending of the word was simplified to fit Moten's strict syllable structure. Moten's phonotactic rules are still in place. It's just that in borrowings, those four letters behave like single consonants whatever their position. The last rule forbids the *ts and *dz sequences, even across syllable boundaries. As I'll show below in the section about morphophonemics, those are always simplified to |s and |z respectively. In a similar fashion, the sequences *lj and *nj never appear in Moten. More surprisingly, neither do the sequences *jl and *jn.
Besides those rules, consonant clusters (whether in onsets or across syllables) are quite free. Plosive + plosive or fricative sequences are considered normal even in syllable onsets, and both s and z can be followed by another consonant (including plosives). Other continuants don't allow such freedom, and actually continuant + continuant sequences are quite limited in onsets (they are very free across syllables). Mostly only those starting with s or z are allowed. In the same way, all consonants can appear in coda (even absolute codas at the end of words), including plosives. Continuants are preferred in that situation though.
Moten is relatively free in allowing word compounding, and inflects using a combination of prefixes, infixes and suffixes. This means that it is common for words and/or affixes to combine in ways that result in forbidden clusters according to phonotactic rules. This section gives an overview of the possible ways those clusters are resolved, although the rules are complex and not always regularly followed (when it comes to compounds. Inflections always follow regular morphophonemic rules).
The simplest case is when two vowels meet. As indicated in the section above, a, e and o can happily coexist (although the sequence ae is sometimes simplified to e). The other rules are:
- When two identical vowels meet, they are simplified into one.
- When i meets another vowel, it becomes j. That j might then interact with neighbouring consonants.
- When u meets another vowel, j is inserted to break the sequence.
Consonants are more complicated, as there are many more possible cases of disallowed clusters. Nevertheless, the number of rules governing the treatment of disallowed clusters is still relatively limited.
There are two overarching rules that affect consonantal clusters before any other rule is applied:
- Consecutive consonants must be all voiced or all voiceless (keeping in mind that the nasals, laterals and approximant are voice-neutral). When consonants of different voicing meet, at least one will change to meet the other consonant's voicing. Unfortunately, there doesn't seem to be a single rule that states which consonant should change in which case. There are some tendencies, like the fact that s and z will always be the ones to change voicing when meeting another consonant, while f and v, on the other hand, always impose their voicing to neighbouring consonants. But there doesn't seem to be any such rules for other consonants.
- Two identical consonants (possibly after voicing agreement) will merge in one.
Next are a series of rules for cluster simplification and breaking. Note that those rules don't have any inherent order. They happen more or less simultaneously or in different orders depending on the context:
- The sequences *ts and *dz become respectively |s and |z. In a similar manner, the sequences *lj and *jl become |l, while the sequences *nj and *jn become |n.
- The consonants |l, |n, |s and |z are simplified to l, n, s and z respectively, when in a position where they are disallowed. This rule has exceptions though. In particular, when next to s or z, it's not |s and |z that are simplified, but rather s or z that disappear. Also, when |l and |n appear next to a continuant (other than j, which disappears) and simplifying them would still result in a disallowed cluster (for instance two consonants in coda position), they are not simplified. Rather, an u is inserted to break up the cluster.
- If, despite simplifications, impossible clusters still appear, those are broken by an epenthetic u.
This last rule has a lexical exception. In Moten some roots break the phonotactics of the language by ending with two consonants. Those roots are used as is for inflection and compounding, but rather than using an epenthetic vowel, disallowed clusters involving this coda cluster are resolved by deleting the last consonant of the root. This last consonant is fragile, and will tend to disappear as well when forming a cluster onset in a following syllable, unless it was simplified in another way.
The most typical example of such a root is the numeral vel(d): five. In the nominative, it will surface as vel, while its genitive is veldi. In compounds, in some cases the final (d) is present (even if simplified): vel|ziza: the fifth of the month (vel(d): five + siza: day). In others it's absent: velmune: May (vel(d): five + mune: month). In some cases, it's absent, but its influence is present in voicing effects: velbele: five minutes (vel(d): five + pele: minute, used when telling the time).
Such roots are relatively uncommon, so I will always make them obvious when citing them by putting their last consonant in parentheses, as I did here.
Stress and Pitch
Stress in Moten is usually not distinctive, and very weak in any case. All syllables in a phrase are pronounced equal in time, strength and pitch, except the first syllable of the phrasal head, which receives a very weak high-pitch stress. This stress pattern helps identify the head of a phrase, but it is usually superfluous. For this reason, stress is not indicated in writing.
OK, I believe this will do for now. Sorry for the relatively dry post. I promise the next ones will be more lively, including more examples of the Moten language. Hopefully you still enjoyed it. If you have questions (whatever they are), don't hesitate to ask them in the comments. I will try to answer them to my best ability.
Next post will be about nominal morphology and syntax. I will present noun declension, pronouns, and noun phrase formation. I will also discuss the meaning of the various noun declensions, which can be surprising at first. So stay tuned!