Strata awake, asleep and dreaming

I am publishing the endochrony manifesto in a bilingual edition: fourteen theses, Spanish and English, seven pages.

The lecture No language is older than another already set out the central argument — that a language’s age does not exist, that three times are constantly confused with one another, and that Bergsland and Vogt demonstrated it in 1962 with Norwegian and Icelandic. I will not repeat it here.

What follows are the theses the lecture did not have room to develop, and which are, I think, the most productive.

No language has an age. Each component has its own.

Four mechanisms, and they are not four species of one genus

Diachronic heterogeneity is not preserved by accident. Four mechanisms keep it available, and they operate at different levels — what joins them is their function, not their nature:

Suppletion. Distinct roots fused into one paradigm. Was inside be. It is a relation between the cells of a paradigm.

Conditioned allomorphy. One root split in two by a sound change: kona against kvenna. It is morphophonological.

Paradigmatic recruitment. The rule forming the superlative operates on a different historical stratum from the one supplying the positive form. The superlative of Spanish pobre is not formed on pobre, but on its Latin ancestor brought back a thousand years later. That is paupérrimo. It is lexical and historical.

Positional fossil. The stratum that survives where nobody analyses it any more. It is a condition of opacity.

Confusing them is the sure way to measure none of them.

Strata are awake, asleep or dreaming

And they are not detected by testimony. They are detected by effects.

Asleep. Spanish usted comes from vuestra merced, “your grace”. Nobody analyses it — yet it governs third person: ¿vive usted lejos?, not vives. Agreement still obeys, today, a noun phrase that disappeared centuries ago.

The speaker does not know it; the grammar does.

Dreaming. Vagabundo, from Latin vagabundus, is reanalysed in the fifteenth century as vaga + mundo, “wanders the world”. Inherited material recombined into a figure history did not supply.

And the direction is not random. Sometimes the dream wins: nigromancia — from necromantia, divination by the dead — was reread through niger, “black”, and erased the original.

Dreaming is not remembering. It is recombining.

The yesterdays occupy positions, and are not distributed at random

This is the thesis that interests me most, because it is the one that can be falsified.

  • The past slot: was, fui
  • The superlative: paupérrimo
  • The genitive plural and compounds: kvenna, kvenmaður
  • Surnames: Fernández
  • Fixed formulas and insults: hideputa

The old withdraws to the periphery and yields the centre — the present, the nominative, the neutral register.

If that ordering repeats across languages, then language has a rhythm: a regular way of distributing its ages across the slots of its own structure. If it does not repeat, the thesis falls. It is written so that it can be checked.

The myth does not live on ignorance alone

Icelandic reads as though it had not changed, but it owes that in large part to an orthography that stood still while pronunciation moved. The text looks ancient. The spoken language, considerably less so.

Concrete artefacts manufacture it: conservative spelling, literary prestige, the date of the first manuscript.

What endochrony does not say

This goes into the manifesto because it is where it gets misread.

It does not say the comparative method fails. It works. What fails is converting quantity of change into quantity of time.

It does not say all languages change alike. It says the opposite, and for identifiable causes — community size, network density, contact, and above all adults learning the language from outside, as Trudgill’s sociolinguistic typology predicts.

It does not say all linguistic systems appeared at once. Creoles and emerging sign languages have real dates. It says that a community’s depth is not read off the date of its first document — and that neither that date nor that depth says anything about how time is distributed inside.

And a declaration of debt

The term is a coinage. The apparatus is not.

Endochrony assembles three existing pieces: Jakobson’s permanently dynamic synchrony — stated ninety years ago — Trudgill’s sociolinguistic typology, and diachronic stratification. Three bodies of work that had never been aimed together at the myth of linguistic age.

What I propose measuring in place of an age is an endochronic profile: not a number but a vector — one rate for phonology, another for morphology, another for syntax, another for the lexicon, each relative to a declared comparison.

“Icelandic is archaic” then stops being a slogan and becomes a profile: conservative in morphology, considerably less so in phonology, with a conservative orthography on top exaggerating both.

The manifesto: Endocronía / Endochrony: the time inside language (bilingual PDF, ES/EN)

The lecture: Ninguna lengua es más antigua que otra (PDF in Spanish, 40 minutes)