So this has been intriguing me for years:

In 'Perspectives on the Quechua-Aymara Contact Relationship and the Lexicon and Phonology of Pre-Proto-Aymara', Nicholas Emlen mentions, citing Adelaar (1986) and Parker (1969), that "many Quechua roots appear to be lexicalized from morphologically complex constructions", e.g. this fascinating example:

miči- ‘to pasture’ miku- ‘to eat’

Which, hypothetically, could have come from an ancient root či and a reflexive suffix -ku, respectively, having been appended to an ancient monosyllabic verb root *mi- meaning something like "to eat" (?)

He also mentions, again citing Adelaar (2009), that several words beginning with [wa] that have to do with hanging, tying, or pulling. All of which I cannot help but associate with monosyllabic tonal languages where reliance on tone allows for that kind of short morpheme.

So my question is twofold:

  1. Could the precursor to Pre-Proto-Quechua have belonged to a family of tonal languages with mostly monosyllabic words? That then underwent agglutination and subsequent de-tonalization due to contact with Pre-Proto-Aymara or some other cause?

  2. Other than conjecture as part of a larger inquiry concerning contact between Quechuan and Aymaran over millennia, does anyone know of any research currently underway to use—I don't know, maybe tonal patterns in Quechuan clauses, or other lexical evidence—to try to wrestle some clues from the data as to the possibility of the precursor to Pre-Proto-Quechua having been a monosyllabic (possibly tonal) language?

Sorry for the lengthy question, but it's been bugging me ever since I stumbled on that paragraph years ago. Can't help wondering what relationships to other language families in the Americas that line of inquiry could reveal.

Thank you, as always, for your time and patience.


Works Cited:

EMLEN, NICHOLAS Q. 2017. Perspectives on the Quechua-Aymara contact relationship and the lexicon and phonology of Pre-Proto-Aymara. Leiden University,

ADELAAR, WILLEM F. H. 2009. Modeling convergence: Towards a reconstruction of the history of Quechuan–Aymaran interaction. Leiden University,

ADELAAR, WILLEM F. H. 1986. La relación quechua-aru: Perspectivas para la separación del léxico. Revista Andina, 4 (2) (1986), pp. 379-426.

PARKER, GARY J. 1969. Comparative Quechua phonology and grammar III: Proto- Quechua lexicon. University of Hawaii Working Papers in Linguistics 1:1–61.

  • 6
    Certainly, it could. Any Proto-language could in theory have evolved that way. Whether there’s any reason to assume it did is another matter, one that I don’t know enough about Quechuan or Aymaran to have any qualified input on. But I can say that using modern Quechuan tonal patterns to reconstruct a putative monosyllabic tonal Pre-Proto-Quechuan is almost certainly a dead end. Tones are generally far too unstable for that. Feb 20 at 10:44
  • Thank you very much for taking the time to answer. It makes sense that tones would be unstable. I'm not too familiar with their stability in language evolution, but now that you mention the issue—supposedly Mandarin tones are a relatively late innovation, and Swahili seems to have lots its tones, so... excellent insight; thank you. Now wondering if it's a lost cause to even try to delve into that topic, or if there are plausible ways we could explore such a distant horizon to at least some level of detail. Guess for now it remains just an interesting conjecture.
    – rcgy
    Feb 21 at 12:30

2 Answers 2


This falls under the historical reconstruction of tone - not an easy job, and even more difficult if there are no contemporaneous records of the language or of discourse about the language that have survived; and even more fatally for the comparative method, if none of the descendant languages retain the feature.

A comparison can be drawn with the reconstruction of nasalisation for final -m in Latin: not something that survived as a measurable reflex into any of the 21st century Romance languages. Contrast this with the issue of vowel length in Latin, which survived in no 21st century descendant but is easily deduced (otherwise we would have to posit a super-large vowel inventory, and would still need to explain Sardinian and South Lucanian of Basilicata).

Proto-Indo-European has been reconstructed with pitch accent, based mostly on the (mis)matching of pitch accents between Vedic Sanskrit, Ancient Greek, and the reconstruction of Proto-Germanic stress (especially with Gothic involved), which affected the vowel ablaut.

Back to reconstructing tone without 'documentary' evidence: Proto-Bantu is reconstructed as a two-tone system. This is despite certain modern Bantu languages (e.g. Swahili, Tooro) being toneless, and also despite different numbers of tones in modern languages [with some modern Bantu languages having rising, falling and even rise-fall tones], and also despite the large split between Northwest / Forest Bantu having the mirror-image tonal reflexes to Central / Savannah Bantu (i.e. where one has high, the other has low). In short, the Proto-Bantu tone pattern is believed to have been most similar to certain Guthrie Zone B/C languages, which appear to be transitional, e.g. Lingala, Kele, Ngombe.

In the development and spread of the Bantu languages, it is possible to see tone spreading, tone anticipation and downstep emerge, possibly out of syllables being lost through time in some languages. Some of these are still present synchronically in many Bantu languages, e.g. Meeussen's rule.

Proto-Japonic has been reconstructed as having some kind of pitch accent or tonal system, where the vast majority of modern Japonic languages have some sort of pitch accent system, but the number of classes is under debate, and the nature of the difference even murkier.

Returning to the Pre-Proto-Quechua hypothesis; in the absence of any tones in modern daughter languages, we would need to find some sort of alternation, most likely in vowels, that would necessitate this sort of explanation. Monosyllabic morphemes, although globally generally linked to tonal languages, may be a clue, but is inconclusive on its own. It is true that there are interesting stress patterns in Quechuan languages; would this be something to try to reconstruct?


The available evidence only supports the hypothesis that the proto-language (if there is one) has penultimate stress. That is how all of the extant languages work, and there is no "lurking mystery" suggesting some phonemic distinction that was lost, analogous to Indo-European laryngeals. The motivation for the question is that there are

several words beginning with [wa] that have to do with hanging, tying, or pulling. All of which I cannot help but associate with monosyllabic tonal languages where reliance on tone allows for that kind of short morpheme.

There is no special prohibition against short morphemes that is alleviated by having tone, indeed there is a better generalization that polysyllabic morphemes tend to be limited to roots. There are two connections between average syllable count and tone. One is a social connection, that Chinese is typically said to have only monosyllabic morphemes, and Chinese is tonal. There is a similar belief about Vietnamese. However, most tone languages have polysyllabic morphemes.

The second is a possible causal historical connection (realized in a number of Benue-Congo languages in the western areas – Gabon, Cameroon, Nigeria), where an increase in the number of tones can be attributed to reduction of words to monosyllables, by deletion of the final vowel. This is well-documented in the Grassfields Bantu languages, and neighboring languages.

The initial premise that there is something unusual needing explanation, in Quechua (with or without Aymara) is incorrect.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.