Dialects and Daughter-Languages

Sapetni founded the Dha-Patu Academy with the new language's first adherents while it was still a work in progress. Once the initial work was done, the Academy's purpose was to preserve the purity of the language and to make any necessary changes; in particular, it was to publish and periodically update the official dictionary of Dha-Patu.

The language spread rapidly; but even its enthusiasts were concerned about the real and perceived defects of their language. Dha-Patu had indeed all the virtues that its inventor had intended for it: simplicity, logic, and consistency; but Sapetni had neglected one aspect of language, and so left it imperfect: it lacked beauty. All its words were, after all, originally random collections of vocables, often rather uncouth. Some editing had already been done, of necessity, eliminating completely unpronounceable combinations, but those that remained were often awkward and difficult. Many wanted to reshape the language generally, making it more mellifluous and esthetically satisfying. In addition to these relatively modest proposals, however, there were already hints of more radical departures to come.

Sapetni, an attentive student of glottopoietic history, had anticipated this: in the foreword to A Project for a New Language, the creator invited the readers to adapt the language as they saw fit, for "change, growth and diversity are the surest signs of flourishing life." Privately, Sapetni expected the freedom to play with the language to attract users to it, while the various artificial dialects spawned by this freedom would effectively cancel each other out.

As various proposals for reforming the language were vetted, the most prominent innovators came together, drew up a joint "Petition of Reforms," and submitted it to the Academy. Being properly concerned that their reforms should follow or simulate, as far as possible, the natural evolution of language, the Petition was modest enough. Phonological reform was, of course, at the top; but it also argued, for instance, that the most commonly used terms must also be the easiest to use, and abbreviated by design, anticipating their abbreviation by spontaneous evolution. To some extent, Sapetni had already reasoned thus: hence the class of particles, distinguished from the radicals. But the reformers wanted to develop this much farther, extending the principle to markers of number, case, tense, aspect, etc.

Even at this early and moderate stage, the innovators were not unanimous on all the changes they advocated. For example, there were, broadly speaking, two approaches to the increase of euphony: to break up awkward clusters with inserted elements (e.g., glides and epenthetic vowels), or to reduce them by eliminating or merging the sounds. Thus the original language diverged into two types of dialects, the abbreviating and the elongating. In some cases, the differences were radical indeed: for instance, ngoe, originally pronounced [ngoe], became at one extreme g, and at the other extreme nogowe.

Observing these things, and being jealous of the language's unity, the Academy rejected all the proposed innovations. In a fit of perversity, then, the innovators went even further, adding ever more sounds and letters, modifying the vocabulary, and even tampering with the grammar, spawning dozens of new, artificial dialects. The ringleader of the innovators, Ostamage by name, proposed a series of reforms so radical that they created a whole new language.

The Petitioners (as they came to be called) failed to act in concert past the signing of the Petition itself. Some radicals (Ostamage in particular) tried to found their own Academies to rival the Academy, but almost all of them either lost interest themselves, or failed to attract much of a following; the few who persisted, only succeeded in isolating themselves in separate languages. The moderates, wishing (quite as much as the conservatives) to keep the language united, attempted to work within the Academy, where as often as not they worked at cross-purposes.

The struggles for control of the Academy resulted, in the end, in a tacit compromise: the written form of the language was simply left as it had been -- which meant that it slowly but steadily diverged from the spoken language. As the "War of the Words" abated, deliberate projects for reform gave way to spontaneous mutation, as the new language was adapted to daily use. Long usage gradually softened the rough edges, satisfying most of the demand for change.

Gradually, seven major dialects (or groups of related dialects) emerged. There were five regional dialect-groups: Central, Northern, Southern, Eastern, and Western. Thomavmic, the speech of the Thmevme or Thomavm, was scattered in many small enclaves. Antipolitan (or Antipodean) was concentrated in the city of Antipolis and its hinterland.

Of these dialects, two developed into separate languages: Thomavmic and Antipolitan. Because Antipolis was the staging-ground for off-planet expansion, this dialect spread throughout the extraplanetary colonies, becoming their Common Speech (Pattungafat), also called Pattudha. Most of Freehold, however, stuck to the old speech and its variants; Antipolis and its vicinity always remained a distinct linguistic enclave. The other five continued to influence each other reciprocally, forming a language called Freeholder. The traits shared by these five dialects are collectively known as "Common Freeholder" (abbreviated CF).


The Central dialect, in the city of Stronghold and its vicinity, became the standard of the Freeholder language, for both political and geographical reasons. The character of the Stronghold dialect was determined by the fact that the Academicians were the most influential speakers and writers in the city.

In reaction to colloquial innovations, which tended to blend, reduce, or drop sounds, the Academicians and their supporters made a point of enunciating their words with pedantic precision, particularly stressing the distinctness of juxtaposed vowels. This had the unforeseen effect of introducing hypercorrective innovations, as some inserted semivocalic glides between the vowels, and others a glottal stop, which, softened to an aspirate h, ultimately became normal in the city, and as characteristic of Stronghold as the Cockney lost h was of London.

At the same time, however, the old pronunciations survived in colloquial parlance: the hypercorrections were introduced only in formal speech. This diglossy led to semantic divergence, as the colloquial pronunciations took on colloquial connotations (cf. ohene 'bad, adverse', one 'lousy, crummy'). In other cases, the Academic pronunciation took on more abstract connotations, while its colloquial doublet was more concrete (cf. sihol 'object', sjol 'thing'; mohet 'sequence', mot 'chain'). This form of derivation is called viring (< [virŋ], forming a a doublet with vng, which retained the old sense of an architectural structure) and became very productive and influential. The high forms of words were adopted directly from Stronghold by the provincial dialects, alongside their divergent colloquial forms.

The aspirate h also entered colloquial speech, though only in sandhi. When the same vowel is reduplicated, an aspirate h is inserted: e.g. dhi itun > dhihitun. Likewise, h is inserted between a vowel and a diphthong, or between two diphthongs: e.g. dhi ajre > dhihajre. When two vowels fall together between a derivational particle and a radical, they are pronounced as a simple diphthong: e.g. dhe itun > dhejtun.

In the Stronghold dialect (and thereafter Common Freeholder and Thomavmic) the weak vowel, represented in standard Dha-Patu with an apostrophe, is realized as a schwa, written (d'nar ~ dnar, shob'd ~ shobd). In colloquial speech, however, the weak vowel and adjoining syllable would be dropped, with similar semantic divergence (cf. shobd 'person', shob 'guy, bloke').

The same often happened with final vowels. Intervocalic consonants were doubled to reinforce them (cf. kaffe 'coffee', kaf 'java, joe'). With final consonant-clusters, similarly, the colloquial pronunciation would drop the final consonant, while the formal pronunciation would reinforce it with a paragogic -: e.g. thint > thin(t).

Whenever the normal processes of composition and derivation produce the sequence VCVCV, the central vowel is elided: orus igen > orsigen. However, it is preserved when two consonants that are the same or very similar would be conjoined (e.g. zhazu-su), though in colloquial speech the whole middle syllable is dropped (zhasu).

In Stronghold, the Isthmus, and along the shores of the North and South Seas, the vowel-clusters were levelled to three basic diphthongs: V, Vj, Vw. The diphthong a was subsequently reduced to aa [a:].

taer > taar, laim > lajm, saop > saap, thaun > thawn
deaz > dez, dheil > dhejl, neof > nef, beur > bewr
dhoar > dhor, moet > mot, zhoin > zhojn, goun > gown
liat > lit, krie > kri, riod > rid, bliu > bliw
luadh > ludh, ruek > ruk, stui > stuj, ruon > run

However, in all seven dialects iV > jV and uV > wV, except after liquids or consonant-clusters.

The Northern and Southern dialect-zones are defined by the two major seas of Freehold: communications by water form united dialects on their respective shores and isles. They are closest dialectically as well as geographically to Stronghold.

By the seas, intervocalic sandhi is pronounced distinctly from Strongholder. In the North, i and e are followed by the semivowel j-, u and o by w-; in the South, the same pattern is followed, but in reverse order -- i and e preceded by j-, etc. In both dialects, the combinations aV or Va (respectively) are pronounced with h.

These two dialects were peculiarly affected by the social and cultural importance of Latin and Greek in the port-cities of Ostia and Piraeus, respectively. The dialect of Ostia and the North Sea was distinguished by the wholesale adoption of Latin words, while that of Piraeus and the South Sea adopted Greek ones. (The Central dialect relied heavily on Anglic, while the Eastern and Western were more creative.) They also borrowed basic parts of speech, such as conjunctions (et, kai) and non-human pronouns (id, afto). Moreover, their borrowings were more archaic, or purist, than borrowings in the standard language.

The Eastern and Western zones accreted innovations as they advanced around the planet, isogloss after isogloss, forming two dialectical continua that diverged to the point of mutual unintelligibility. Both are, in most cases, abbreviating dialects, and the one tends to change or drop what the other kept, since the Easterners stress initial syllables, whereas the Westerners stress final syllables. For instance, the Eastern dialect drops final vowels (e.g. bune > bun) and the Western dialect drops initial vowels (e.g. ezek > zek).

In the Eastern dialect, all vowels in unstressed syllables are levelled to (e.g. ensi > ens, ezek > ezk), except vowels before final sonorants, which are totally lost, the sonorants being vocalized (e.g. itun > itn). In the Western dialect, initial syllables with sonorants are reduced the same way (e.g. ensi > nsi).

Likewise, in the East diphthongs are falling, and in the West rising. In the East, the combination aV either remains diphthongized or merges into long or hybrid vowels:

In the West, aV > V > 0V, while e- > j- and o- > w-:

After liquids and consonant-clusters, Westerners pronounce the vowels distinctly, with glides (e.g. lijat, luwadh).

In Common Freeholder, the agentives were reanalyzed as modifiers of the noun, and merged with it as case-marking suffixes: for the subject of a transitive verb, the subject of an intransitive or reflexive verb, and the direct object of a transitive verb (a peculiar hybrid of nominative- and ergative-type declensions). The genitive and dative are marked by -0. This allowed the order of elements in a sentence to be changed, though SVO remained standard, with alterations only for emphasis.

The original pronouns became annexed to the verb, with the pattern S-verbal root-DO-IO. The subjects of passive verbs would be suffixed to the verbal root. When inanimate objects are to be indicated, demonstrative particles were used, until they were levelled to s(i), derived from su under the influence of siol.

The verb, then, might form a complete sentence by itself: e.g. japrashjo 'I see him'. The pronominal markers might or might not be used if their referents are explicit. At least one marker is required to distinguish the verb from an adjective; for emphasis, the verb could summarize the entire sentence: dhashobdta (jo)dringsi 'person (himself)-hits-it'.

Verbs are nominalized by periphrastic constructions with srot: e.g. dhipatsrot 'speech, to speak'. Verbal modifiers are constructed likewise: e.g. patsrot 'speaking'.

A whole new series of personal pronouns was formed by compounding the pronominal markers with demonstratives, making the indirect phrases 'this person' (for 'I'), 'that person' ('you'), etc. The vowels in the pronominal markers were levelled to : hence jpi, jko, jsu, wpi (exclusive), wte (inclusive), wko, wsu. Because pronominal markers are inherent in the verb, the separate personal pronouns are only used for emphasis. Possession is marked with pronominal particles, e.g. dhasoghja 'my hand'.


The Thmevme or Thomavm were the product of a small faction of Neuters who decided that their creation had been a mistake, and re-engineered themselves into male and female. Their ethnonym simply means 'the Sexual People' (as distinct from Dha-Thome, 'the People', as the Neuters call themselves). Physically, they still resemble the Neuters in many ways: slight of build, with high foreheads and small chins. The women have small breasts and slim hips, the men are beardless; both sexes completely lack hair on the body.

The dialect of the Thomevme was spoken, more or less uniformly, in all their scattered enclaves on Freehold and the colonies. They were the only Dha-Patu speakers who emigrated in groups large and cohesive enough to form their own speech-communities.

As the Thomevme themselves took on sex, their language took on gender. The primary innovation was the generalization of Latin and Greek gender-markers, which in standard Dha-Patu are used only to form proper names: e.g. Karolus/Karola, Aleksandros/Aleksandra. In Thomavmic, they are used to mark nouns of natural gender: not only male and female persons and animals (e.g. shobdos/shobda 'man/woman') but male and female attributes and things pertaining to males and females (e.g. adbos/adba 'men's/women's association, moiety'), which is a very productive derivational process. The choice of the final vowel was determined by harmony with the penultimate vowel in the noun, based on the height of the vowels: cf. asra 'female animal', pige 'woman's buttocks'.

The next stage of development was the loss of final -s, which resulted in the complete confusion of gender-marking suffixes with pre-existing, semantically insignificant final vowels. By process of elimination, -i then acquired a neuter connotation. This had two consequences: all final vowels > -0 or -i to signify common gender, or were altered (or added) to signify the masculine or feminine genders -- even in adjectives, which now agreed in gender with nouns. Final vowels (except -i and -) were eliminated from roots when used as verbs and adverbs.

Thomavmic, generally speaking, follows the Freehold dialects phonetically, but the Common Speech morphologically. That is to say, the forms of its words usually follow the Stronghold standard (except, of course, when adapted to the new gender-system), but like the Common Speech it reduces both articles and agentives to proclitics. However, the development of final gender-markers precluded the accretion of agglutinative suffixes.

The cradle of the Thomavmic dialect was Erotopolis on the Isthmus, which followed the Central dialect in the development of its diphthongs, except that final -a remained in Thomavmic (e.g. idao > C. idaa, Th. ida). Final vowels after consonant-clusters > i (e.g. adbo > adbi) unless the second consonant is a sonorant, in which case it is vocalized (e.g. asru > asr).

The vowels in the derivational particles dha and ta are elided before vowels in radicals (e.g. dhuraj < dhe urai). However, dhe and dhi are reinforced by j- (e.g. dhejuraj), while to and tu are reinforced by w- (e.g. towuraj).

Thomavmic borrowed a comprehensively gender-marking third-person singular pronoun-system from Anglic: hi 'he', shi 'she', dhej 'he or she', t. Only Neuters are referred to as jo. The third person plural is formed with -z ~ s.

The lexis of Thomavmic is naturally distinctive for its elaborate sexual vocabulary, formed by both derivation and composition: e.g. dhisune 'breastage', i.e. display of the breast(s), including dhisunnefa 'cleavage' < suni neof. Thomavmic compounds were often adopted into Pattudha, supplementing the native vocabulary (which was exhausted by the three words avme, deaz, and itun) and anatomical terms derived from Latin and Greek.


The city of Antipolis was founded at the antipodes of Stronghold, where the eastward and westward streams of migration met and mingled. As speakers of widely divergent dialects relearned how to talk to each other, local interest in language-reform was quickened, and the new city attracted planetwide attention from reformers whose aspirations for had lain dormant since the War of the Words. Having had time to reflect on, discuss, and learn from their experiences, many of the old Petitioners (joined by other, younger reformers) now gathered in Antipolis and founded an "Anti-Academy" to establish a new, recast form of the language. A minimum, compromise program was worked out and finally agreed on: a synthesis of the most popular sound-changes, and spelling-reform to reflect those changes. The dialect caught on, deliberately cultivated by the Antipolitans as a counterpoint to Stronghold and the Academy.

Where necessary, initial and final consonant-clusters were broken up by reduplicating the next vowel, with the accent on the original vowel (e.g. d'nar ~ danr, shob'd ~ shobod).

All diphthongs were reduced to single sounds, either long or of intermediate quality:

Intervocalic consonants were doubled: e.g. bune > bunne, ezek > ezzek.

Medial consonant-clusters were assimilated or dissimilated, as appropriate: e.g. apzi > apsi, ithso > itso.

Whenever juxtaposed with gutturals, l > r: e.g. gluv > gruv, ighla > ighra, zhalg > zharg, shakl > shakr. Before liquids, dental stops > fricatives: e.g. dring > dhring, tlur > thlur, itle > ithle (CF dring, tlur, itle).

The affricate category was reinforced by the assimilation of sibilants after nasals (e.g. jans > janc).

Initial vocalized sonorants > VC pairs: e.g. ntok > ontok.

Whenever the radical begins with a vowel, the vowel of the derivational particle is elided, so that the class of the noun or voice of the verb has to be inferred from context. This elision may be avoided by reduplicating the proclitic: e.g. dh'urre ~ dhedh'urre. This process is used when forming the definite article.

Once again, the pronominal system proved the least stable. At the outset, it was replaced by one borrowed from Esperanto: mi 'I, me, my'; ni 'we, us, our'; ti 'you(r) [singular]'; vi 'you(r) [plural]; oni 'one', si [reflexive]. After some experimentation, new third-person pronouns were formed from su + shobod, sjol: subod, sujol, compounded with -tor [dual; idiomatically, subottor came to mean a couple, i.e. a man and woman], -nar [paucal], -zek [plural].

The name Pattudha for the later development of the Pattungafat is derived from the suffixation of dha and dhe to proper nouns (i.e., names of things or places) to serve as proper (as opposed to common) articles. The original pronouns are similarly used with nouns and phrases that serve as the names of persons, i.e. as personal articles: -jo and -wo are roughly translatable as 'he/they who ...'. These referential affixes change according to whether the name belongs to the person(s) speaking, addressed, or spoken of: thus -je and -we are equivalent to a vocative case.

The agglutinative formation of suffixes, derived either from particles or enclitics, was extremely productive in all parts of speech. For instance, the noun now has four intrinsic numbers: singular, dual, paucal, and plural, the last three formed by suffixing -tor 'two', -nar 'few', and -zek 'many', respectively. Enclitic -(st)i is used with dh(e)- or -wo to signify 'one member of a group'.

Enclitic forms of the words for 'male' and 'female' (-tun and -dez) were directly affixed to nouns and pronouns to indicate their sex, when appropriate. Nonhuman gender (plants, animals, inanimate objects) is marked by -sjol 'thing'. The use of gender markers never became completely consistent (nor as thoroughgoing as in Thomavmic). The new distinction between masculine and feminine, in effect, created subcategories of the common gender. There never was any parallel way of specifying a neuter person.

The pronominal system has the same four genders and four numbers as the noun, but gender is only marked for the third person. The enclitic -shob, derived from shob'd (forming a doublet with shobod), is the common-gender marker in the first and second person singular; -tun and -dez are used only in the third person singular, while joshob is used for Neuters. The non-singular numbers are marked with with the same affixes as nouns.

The verb has three tenses: past, -tjuth; present, -0; and future, -re. There are three aspects: inceptive, -(e)vne; progressive, -jus; and perfective, -ghub. The tense markers always precede the aspect markers, if any. Other tenses and aspects, mood, and other conditions of the verb are still indicated (optionally) by various adverbial modifiers, or by context.


Table: Numbers in Patic Languages

# Common Freeholder Thomavmic Pattudha
1 i(sti) isti ~ istu ~ iste isti
2 (p)tor (p)tor ~ toro ~ tora potr
3 lihen ~ lin lin ~ linu ~ line ln
4 umno umn ~ umnu ~ umne ummo
5 tlur tlur ~ tluru ~ tlure thlur
6 epud ~ ept epud ~ epdo ~ epda eppud
7 ur(i)p urip ~ urpu ~ urpe urrip
8 sahop ~ sp sp ~ sapo ~ sapa sp
9 tihog ~ tjog tjog ~ tjogo ~ tjoga tsog
10 afohi ~ iho afoj ~ afojo ~ afoja aff
11 ihohi afoj-isti aff-isti
20 toro afojptor afftor
21 torohi afojptor-isti afftor-isti
100 shakl ~ ihoho shakl shakr
101 shakli shakl-isti shakr-sti
200 toroho shaklptor shakrtor
201 torohi shaklptor-isti shakrtor-sti

Appendix: Names and Naming