Background

Occam’s razor … states that the explanation of any phenomenon should make as few assumptions as possible, eliminating those that make no difference in the observable predictions of the explanatory hypothesis or theory. The principle is often expressed in Latin as the lex parsimoniae (‘law of parsimony’ or ‘law of succinctness’): ‘entia non sunt multiplicanda praeter necessitatem’, or ‘entities should not be multiplied beyond necessity’.

This is often paraphrased as ‘All things being equal, the simplest solution is the best.’ In other words, when multiple competing theories are equal in other respects, the principle recommends selecting the theory that introduces the fewest assumptions and postulates the fewest entities.

Occam’s razor

As you have noticed, Lemizh is a conlang or constructed language. It is an attempt to use Occam’s razor for creating a grammar, not because this should result in a better language, but because that’s how my brain works.

This means that the ‘observable predictions’ – to stay in the diction of the quoted Wikipedia entry – are the ideas we can represent in speech; the ‘explanatory theory’ is of course the grammar, used to arrange these ideas in a fashion typical of this language; and the ‘assumptions’ of which we should need as few as possible are the rules that constitute that grammar. Lemizh is the best solution I could find (provided it is a solution). To get an idea of its simplicity, have a look at the page Summing up, unless you have already been there. But don’t expect to understand much of it before you have read the tutorial.

The regularity of the grammar seems to imply that Lemizh is an engineered language but, as noted on the history page, it might as well be the result of a series of strange coincidences in a natural language.

Epiphenomena

What fascinates me about Lemizh grammar is that it reproduces a lot of the behaviour of natural science, if on a much smaller scale.

• Just as physics moves further and further away from ‘common sense’, Lemizh has ever been moving away from Indo-European grammar, though at no point I decided on a rule for its strangeness or alienity (except perhaps with the hexadecimal number system). All changes were applied with reluctance – e.g., I resisted a long time to merge verbs and nouns into a single part of speech.

  At one point I had to discard most of the grammar (about the whole body of rules except verbal endings) to solve two minor problems, robbing myself of the possibility of expressing adjectives, participles, genitive attributes, questions, and commands. The modern sentence grammar is based completely on the solution for these two problems. Why does it work?

  If you don’t conclude that I want to compare myself to Einstein, I can tell you that the great physicist explained a simple problem (Why does light always travel at the same speed?) at the cost of a working theory of gravity. It took him ten years to think of a new one.

• I don’t recall making any decisions most of the time; the last one I made sometime during the nineties. The decisions I do remember – the sorting of voiced before voiceless consonants in the alphabet or the use of moræ rather than syllables – had no influence on the overall structure of the language. Usually I am happy to find a single solution that doesn’t contradict anything else. The language compels me to do things – I hadn’t any choice with Identity of action or Completeness of cases, they just forced themselves into existence. Yet the creation of Lemizh is far from mechanistic or reproducible, it is my very own work.

  There’s something to think about, if we’re talking of freedom of choice.

• And then there are, yes, epiphenomena. Concepts such as bracket, inversion or inner agentive are not grammatical rules but applications or consequences of the rules. Some of them have proved more powerful than I imagined. I couldn’t have made up a rule as versatile as inversion, yet there it is. Who invented it?

And the lexicon?

I was persuaded (mainly by Umberto Eco and Jorge Louis Borges) that Occam’s razor is unserviceable for creating a lexicon, and especially for the definition of sememes (units of meaning: ‘tortoise’ [tortoise] or ‘-s’ [plural]). The space of ideas, so to say, is not measureable.

I could make up arbitrary words as in Esperanto or Klingon, but people would ascribe this to my laziness. (At least I would.)

So the only way is to produce a completely anti-Occamian vocabulary: as many rules as possible, with any number of contradictions, cross links and short circuits. Not as we would conceive a scientific theory, but a novel or a symphony. Abundancy is the word, not succinctness.

• Each language group that contributes loanwords will have an ‘impossible’ sound shift. For example, New Troyan consonant clusters are calculated using a multiplicative quaternion group of order 48, with each group element corresponding to a consonant or consonant cluster. It is closely related to the symmetry group of the cube, based on the facts that Troyan is related to Greek, and that the Greek invented geometry.
  Ghean words are created by a pseudorandom number generator that uses English glosses as seeds.

  Such mechanisms are of course impossible in historical linguistics and can only occur in a constructed language. On the other hand, if we see Lemizh as a natural language, they might just be weird coincidences like the regularity of the grammar.

• Every kind of sound shift has an associated semantic shift. This has the triple function of creating (a) more arbitrary and less artificial semantic shifts, (b) connotations, where a semantic shift does not lend itself to inclusion in the word’s primary meaning, and (c) yet more impossibilities or coincidences, whichever you prefer.

• Ideally, every word would have an in-world history, given in the dictionary, as well as a meta-history, to be deduced or suspected.

I am nowhere near these goals. I know.

Examples

• xtÌ. ‘eight’ elides o in Old Lemizh, a shift which translates to ‘locative narrowing’. Now the place for making eights (for want of a place for making eight individuals) is a frozen lake, with associations about low temperatures and hence the North. (The Big Dipper is just a feeble excuse.) Elimination of U in Late Middle Lemizh means ‘instrumental narrowing’ – and as the instrument for making eights on a frozen lake are the ice skates’ blades, we arrive at the number’s ‘edgy quality’.
• flàc. /ɸlʌʒ/ ‘to make/colour blue’ is the wholly regular outcome of PIE *bʰleh₁‑ ‘shine, flash’. No tricks.
• kràt. ‘to hunt’ is one of my favourites. This word, too, is the regular outcome of a PIE root, accidentally resulting in the first syllable of ‘Κρατύλος’, the title of a brilliant wild-goose chase for the origin of words. With the second component ylàs., admittedly, I did cheat.
  And since Epicurus is a cognate, I threw in the phrase ‘the pursuit of happiness’ for good measure.

What verbs does the language need?

Finally, a remark on a problem I frequently encounter while creating words, mainly with nouns. What is the underlying verb? The default for a concrete noun such as drÌzd. ‘chair’ is to introduce a verb ‘to make/build chairs’, resulting in an inner accusative for the noun. But often there are other possibilities: gwrù. ‘knife’ is a tool noun derived with an inner instrumental from a verb meaning ‘to cut with a knife’. I could as well have derived it from a verb ‘to make knifes’. On the other hand, there could have been a verb ‘to sit on a chair’, making ‘chair’ a tool noun or perhaps one with an inner scenic case… Come to think of it, what prevented me from deriving ‘chicken’ with an inner accusative from a verb ‘to eat chicken’, or ‘wolf’ with an inner dative from a verb ‘to dance with wolves’?

The lession is (and this took me quite a while to realise) that this is, so to speak, not a question for Occam but for Borges. Any of these solutions, and many more, would work grammatically, sometimes resulting in simpler, sometimes in more verbose sentences. So, if you ask yourself why ‘leg’ is a tool noun but not ‘arm’, why ‘earth, land’ is the location (loc) of making earthly but ‘sky’ is the content (acc) of sky-making, or why trees are the content of tree-making but fruit and nut trees are the source (nom) of their fruit: often the answer is that I did what I thought the Lemizh would do, and that’s all there is to it.

And so …

He knows that there are in the soul tints more bewildering, more numberless, and more nameless than the colours of an autumn forest; he knows that there are abroad in the world and doing strange and terrible service in it crimes that have never been condemned and virtues that have never been christened. Yet he seriously believes that these things can every one of them, in all their tones and semi-tones, in all their blends and unions, be accurately represented by an arbitrary system of grunts and squeals. He believes that an ordinary civilised stockbroker can really produce out of his own inside noises which denote all the mysteries of memory and all the agonies of desire.

The limits of my language mean the limits of my world.

Thanks