So, this language -- which is no longer pfft but whose new name is in flux -- has brewed on over the years unexposed.  But it now has a handful of words and a program.  The program is to write and deliver a brief description (15 minutes tops) of the language in the language and "film" the presentation with an overprint of the text and translation, then write up the text with additional comments about the internal structure and also the process of creation.  Right now, the first three sentences are tentatively done.

The fiction is that a retired academic is delivering the lecture to explain the development that led to vowellessness and the moves to maintain it.  But, to do so, he must say a bit about the structure of the three languages that feed into this final project and what gets borrowed/ devolves from each.  So background languages with a kitchen sink of features and a naturalistic development to something feasible.  There is also an inherited spelling system which develops along with the various parent languages and continues though being replaced by a more scientific one.  The old spelling helps to explain some phonological developments, since it show earlier forms with intact vowels to work their phonetic magic.  The new system shows the latest results.

What exists so far:

tcc
Taisio 

greetings

[ba taixo]

pccn djjm fnn fllstmmts mm
Poinktsianen giamen fini filostaemits mimen

I am Porquier, an emeritus member of the Academy.
[Pig seller family(subj) member (perf asp) sage tribe (partitive) I
Poink testu (cf tetestu  thht "merchant"). Ian (family, cf staem, stmm, tribe, group generally, vasle vssl  nation gr). Filo(sofo) sage staem "tribe, organization", its partitive [en],  mim I, me -(e)n lost after final nasals.  Family full form ghian qnn
cf Pfft wedding , marriage pepod redup for importance of pod sale, barter see also redup on member for officer, again for president of academy, tribes generally: gegiamen gjjm Councilors and qqdjjm Officers'

Djjm giama,  member [lb djime]. Fnn – fini perfective aspect [lat]

Xhhnsk drr knnh drr mmf  bwwn hnn

Chothanisk der quonute der mimou .borinen iten
This is a lecture about my language , Chothanisk.
[Chothanisk about language about my possessive lecture subj this subj
Chothanisk Chotan a place name + -isk generic adj former [gr khthon] knnh [back to front of tongue, so the whole thing, hence also language] der dative of reference sorta [la de re] f/v –ou partitive genitve en [gr]] borin lecture [en duh!] hh ite this near point

What is missing here is the information that /c/ is esh,  /h/ is theta,  /j/ is ezh, /q/ is gamma, /r/ is retroflex r, /w/ is trill r, /x/ is chi and /y/ is edth.  double letters are syllabic peaks (the alternative of using a dummy vowel here was rejected for aesthetic reasons).

The first line in each case is the phonetic version, the second the old spelling, the thrid the English.  The rest is commentary, still somewhat higgledy-piggledy.

Wanderings IIIb Semantics

This was supposed to be about verbs, which are, after all, said to be the basis of the language.   But my head ran off to other things. Like
Personal Pronouns
The first, nonsingulars, have both inclusive and exclusive forms.

The second has three forms respectful, familiar, dismissive, for addressing those above, those on the same level and those below.  These levels are very precisely defined and take effect immediately: There is a story about an official who heard that a person he had known all his life and who had come up the rank ladder just behind him had just been promoted to his level.  He went to congratulate him and say how happy he was to now be able to use the familiar form with him. His acquaintance informed him that he had just been demoted and his former inferior now took his place -- all said with the dismissive form. The three forms are said to derive (it is possible but hard to prove) from three responses to direct address: judging, chatting, and obeying.  The first person form derives from the verb for speaking.  Nonsingular forms always go the highest rank among the hearers (when in doubt, respectful) and the inclusive non-singulars have the same diversity and rules. Women a children rank a rank below their male guardian, should a man ever speak to one, and nonhuman and inanimate objects (should they need to be addressed) are always dismissive.

The third, derived from a verb that is mainly used to say "there is/are", takes the "gender" mark of its referent, if that is clear, otherwise a bare form, "one", roughly, which is also used impersonally.

The language has about a dozen noun classes, marked by distinctive prefixes. No pattern for the division has ever been worked out, though, in some cases, new words have been fitted into classes with some sort of ad hoc explanation. Prefixes go on the noun and a direct modifier (however remotely placed) and on anaphoric pronouns, both restrictive and non-restrictive, but not on a predicate modifier (as we would say, it is strictly a verb).

Deictic pronouns are based on the verbs for pointing, aiming and hunting and sporadically use "gender" markers for the intended object  (and always do when used adjectivally).

Tenses proper:
A present tense and an anticipation tense, then that pair transported to a remembered time are the basic units.  From each of these four stable points extend vectors forward and backward  as well as spreading out over the time immediately around the point.  There is also a bare form, usually to mark the central events around which the rest are organized (officially), for each point is defined by the occurrence of some event.  In the official line, pastward vectors mark events which are relevant to the present even only in passing, while the pastward displacement of a point marks an event that needs to be developed in dealing with the present event.  Futureward vectors different from anticipation points only in degrees of the speaker's confidence in their eventual occurrence (more for displacement, less for vector).  Some of these forms have little purely tense use, except, perhaps, in indirect discourse, but they are put to other uses as well (contrary to fact and such like subjunctives, optatives, and so on). 

Wanderings III: Semantics a

Much as I have tried to avoid this, I can't much longer (though I can still put off actually picking words for a while).  So, first, what kind of culture is this a language for?  Since we are working on the ancestral language for now, we can assume at least a medieval culture and probably better a BCE one: rural, low mobility, small community, poor (not quite toki pona's Daoist village, but close).  What is there there then?
1. people, children, elders, friends, enemies, neighbors, aliens, relatives (3 or 4 generations, viewed from several directions, and extending outward from the top.  Maybe much more for genealogists), spouses, in-laws, leaders, healers, spiritual practitioners, genealogist, experts.   male and female (not equal linguistically).  village, tribe, clan, family, nuclear family
2. critters: food animals, other produce animals, working animals (not mutually exclusive, of course), house animals. vermin, dangerous animals, pests, mythic critters.  What goes into each category depends on where we locate this language.  fish, birds, amusing small things, game
3. major crops (we are at least in the sedentary agricultural stage), kitchen crops, wild crops. trees and shrubs, fruits and nuts, herds, pretty things.
4.House, fire, hearth, bed, blankets, headrest, smoke hole, door, window (?), cooking pots and utensils (maybe eating, too), fuel, clothing (many items, different for different sorts), footwear (?), latrines, knives. garbage dumps (?)
5. Wealth: flocks, money (?), land holdings, rent, taxes (or in kind service), lending and usury, bribes/administrative fees
6 War, soldier, weapons (swords, spears, bow and arrow, hammer, having tools, sticks and poles and rocks), rape, pillage, massacre, hunt, poach, steal, buy/bargain, shop, sales person, craftsmen (maker of stuff, anyhow)
7.Physical features: mountains, hills, mounds, fields, plains, lakes, rivers, forests, deserts, mud. rain, snow, storm, thunder, lightning, clouds, sun, moon, stars, comets and other weird stuff, roads, passes, paths, fords, boats and the like and their parts and implements, winds (direction, temperature, force), soil, dirt, dust, sand, rocks and stones, ore, metal, mine, smelter (maybe not for everyone, but we are up to at least the bronze age)
8. carpenter and tools, tinker and tools, sewer and tools, butcher and tools (maybe not specialties yet, but something that someone does from time to time) and materials, mason and tools. herdsman and tools,  farmer and tools, forager and tools  (fisherman and tools, sailor and tools -- at least boatman) weaver and tools,
9. Gods, ancestors, ghosts, local spirits, intrusive spirits and folks to deal with them.

Wanderings II: Grammar

Ideally, the grammar should be such as to provide totally free placement of sentence parts -- within reason.That is, none of those formulae with S, O and V, or A and N need be followed, though total separation of related components might seem to be a dirty trick, at least a defiance of Gricean conventions.  That means that nouns (a term of convenience to be defined later) and adjectives (ditto) need a fairly elaborate case structure (ditto, again).  Similarly, but for different reasons, verbs (ditto) will need fairly elaborate conjugational structures (ditto).

So, nouns (and adjectives agree) have an ergative base (this just fascinates me) but add the motion system of "to(ward), from, at" (maybe for time as well?) and maybe a few more. Vocative is the barest form.  Singular, dual, trial, and plural.  Separate declensions for masculine, feminine, inanimate?  Or some other system, e.g. like Swahili (and related languages)?  Definite and indefinite forms (usual rules with usual exceptions)

Verbs have the usual three persons, but first doubled in dual, trial and plural for inclusive and exclusive forms.  Maybe also reflect the "gender" of the subject or the object.  Many explicitly at least trivalent, subject, object other object, though this is covered by noun cases?

Tenses: standard Bull set, with the retros being just the current one with a -na- infix, vectors the same at all bases, with 0-vector different from bare base.  Retrofuture used for contrary-to-fact cases.  Work out differences between vector and displacement and tensors for both.  Sequence of tenses? Imperative is barest form (retrofuture imperative for value judgments, others not much used in recta, but simple future for hortative?).  Full set of aspects: inchoative, initiative, continuative, pausative, completive, perfective, superfective. Different verb classes?  Active, passive, reflexive-mutual (maybe case change here? i.e. both nominative)

More prefixes (and, maybe, infixes)

Adverbs barest form, adjectives agree with head (generally follow when together, but ...)  Both positive and negative degrees.

Postpositions (some with several meanings depending on the case of the noun)

Both relative clauses, different pronouns, and both different from interrogatives.

Pronouns for all persons in verb conjugations, though some similarities across a given person number.

Free word order for prosody, rhetoric, etc. but tend to favor VSO

There, I think that has enough of  my favorite quirks from linguistics to make a good first conlang.

Wanderings I: Phonology

KISS

OK, so just the basics, with enough to build weird stuff on: p, b, m, t, d, n, k, g, s, r, l and the five vowels.

/m/ is marginal, only occurring in roots with /p, b/.  But it remains even when these are removed somewhat, and /n/ can be infixed without changing.

Accent: rising, falling tone, and level (with automatic adjustment from word to word).  Maybe rising on first syllable, falling on third.

Vowels depend, in part on accent, but also enter autonomously with affixes. Only /i, e/ occur with rising tones, only /u, o/ with falling.  If an affix with a different vowel comes to have a moving tone, the vowel moves to the nearest acceptable one: /i/ and /u/ exchange, /e/ and /o/ also.  /a/ goes to /e/ or /o/ as the case may be. 

Roots are three consonant (I said I didn't have a lot of imagination), although there are a few that involve only two or even one.  There will be some restrictions (like the one on /m/), but I haven't worked out what will be most useful at this point (that may depend on how many roots I end up needing)

All of this will get progessively more complicated as this protolanguage is moved toward the modern form I am sneaking up on.  Having some idea of what that is will guide many choices along the way.

In the beginning. ...

After 35 years of working and playing with language creators and improvers, I decided to try my hand at a language of my own, one just for my own amusement.  Of course, being me, I immediately went off with grandiose plans, far more things than I could realistically accomplish.  After a few bouts of reality, got down to the core of what I wanted:
     1. A language without vowels, with the syllabic peaks being always consonantal continuants and no phonemes having allophones in the vocalic triangle (in the paradigm dialect alt least).  This was the amusing parts, since I like the sounds that are produced, including the language name, pft (subject to change as the program progresses)
     2.  A reasonably natural language (phonology aside), with quirks and peculiarities scattered here and there, but, perversely, in regular ways.
     3. Since, at the beginning at least, the temptation for a language creator is to make the language regular across the board, I needed some way to build in the irregularities I wanted.  Since most of the oddities of most languages are a product their histories: phonological shifts, borrowings, underlying languages of conquered people (or conquering), I decide that my language must have a history, first language from which it was descended through the various vicissitudes of language development.  That is, I will build my contemporary by building a nice orderly ancient language, then screwing it up in the ways languages have managed to screw up their ancestors (or, rather, the inverses of all the ways linguists have devised to get nice orderly languages from the muddle of their descendents).  I am already beginning to see that thre fun is going to be in the process, not so much in the product.

Since all my non-field-related linguistics is in Indo-European languages, my pattern for the ancient language is UrIndoGermanisch: roots strings of consonants, surrounded and separated by vowels which pop into place out of somewhere, under the influence of accent, phonetic environment and affixes, broad ranges of cases, many verbal forms, prepositions (maybe post- for variety), derivatives and combination rules.  I expect I'll begin with something a little simpler than UIG, though. in all these things.  The move then will be by staggered waves of phonetic shifts (vowel weakening, palatalization, rhotacism, ...) and grammatical shifts (periphrastic constructions, word order fixing, reanalysis, ...) and environmental influences (meaning changes with culture, borrowings and incompetents, ...)  I will try to abide by the two great (contradictory) rules of the Junggrammatikern "No exceptions to sound change rules" and "Every word has its own history."

Needless to say, all this planning is to avoid actually starting working and, especially, starting to think of vocabulary.  I'll get my start on that from the usual source, what others have used as basic -- probably (because I know them best) aUI, NSM, toki pona and Swadesh -- plus my own more or less a priori ideas of what is basic (partly derived, however, from problems with the above cases). I don't yet have an idea how to assign roots to various concepts (except that the root for "speak" is ppd), but, if nothing occurs to me, there is always randomness.