I'm not sure if this type of thread is allowed, but I've been pondering the possibility that language began as commands rather than expression of arbitrary thoughts. This is based on the observation that in a majority of world languages the imperative form of the verb is a "root" or "defective" form, lacking in inflections or affixes present in non-imperative conjugations.

From this, I suppose that commands could have been the proto-form of language due to the fact that primitive commands need at most a single verb and a single noun - the object of the sentence; as its the only one, this argument would need no marking for case, and no morphosyntactic alignment. As the ability to add an actor to the sentence, the means by which this is accomplished gave rise to the different alignments as well as topic-prominent structure.

Has this idea been explored before? What evidence is there for and against this idea?

  • 1
    Do you have any proof to "in a majority of world languages the imperative form of the verb is a root"? It seems to be a very strong assumption.
    – bytebuster
    Jan 18 '15 at 11:58
  • At the time I was thinking that factoid was from WALS, but apparently it's from Introduction to Typology: The Unity and Diversity of Language, which cites Speech Act Distinctions in Syntax. Jan 18 '15 at 19:40
  • 2
    I'm voting to close this question as off-topic because it's asking for confirmation of a speculative theory based on faulty assumptions.
    – curiousdannii
    Mar 15 '15 at 5:30
  • Written language goes back about 5,000 years. Human language may be 50,000 years old, or as old as our species, or older. So no one knows what the first languages were like. So we don't know that verbs and nouns were the first constituents, or that commands were the first speech acts. Your question contains unwarranted assumptions and is purely speculative. So I'm voting to close it. Sorry. Mar 16 '15 at 2:23

Despite the sceptical comments, what you say is main-stream linguistic thinking. In Indo-European, Semitic, and most other families the second-person singular imperative is identical with the present stem of the verb.

  • 1
    -1: I don't want to look as a serial downvoter (just downvoted your another answer), but this also does not even attempt to answer the original question; it rather looks like a confirmation of what the OP asks for.
    – bytebuster
    Feb 5 '15 at 16:16

I'm not sure where you got the idea that imperatives are generally the 'root' form. See WALS for the distribution of the morphological imperative: http://wals.info/feature/70A#2/19.3/148.2. The vast majority of the language covered have a morphological imperative expressing person. English is the only language in Europe that does not.

However, the idea that 'non-inflected' forms of verbs are somehow the default is really just an artifact of dictionary making. So you could easily hypothesize that commands were the original impetus in the development of language. You just cannot make a morphological claim about it. In general, any such theories are pure story telling so it is not a very fruitful line inquiry.

  • I think you have misunderstood the link. It is about whether languages have different forms for the singular and plural imperative. Most languages in the world do in fact use the bare verbal stem for the 2nd person singular imperative. English is definitely not "the only language in Europe" to do so.
    – fdb
    Jan 18 '15 at 14:34
  • I see your point. I was focusing on the presence of imperative morphology rather than on what the morphology looks like. Still I think it's a stretch to link the morphology to some original form. Look at Slavic languages or American Indian languages for counterexamples. Jan 18 '15 at 15:38
  • You can see this if you look at languages with a very complicated morphology, like Greek or Sanskrit. Greek “phere” “carry (2nd sing. present imperative)” is the present stem of the verb pherō, it is phere + zero. The “imperative morphology” here is the zero suffix.
    – fdb
    Jan 18 '15 at 15:59
  • I understand that - it's the same in many other languages. But the question in this context is whether 'zero suffix' means there is morphology or 'root' as the questionner put it. I lean towards the former. Jan 18 '15 at 16:03
  • At the time I was thinking that factoid was from WALS, but apparently it's from Introduction to Typology: The Unity and Diversity of Language, which cites Speech Act Distinctions in Syntax. Jan 18 '15 at 19:37

The impossibility of addressing the question empirically led to the Linguistic Society of Paris banning such papers. So any answers are going to be very speculative and heavy on the conceptual side. Commands would no doubt have been part of earliest language, but logically prior to that is conceptualization and conventional symbolization, that is, the ability to identify "bear" as distinct from "lion" or "kitten", likewise "eat" versus "feed". In order to have commands of any kind, you have to have reliable words, otherwise hearers won't have a clue what action they are supposed to perform on what object. That, then, would be the original origin of language, IMO.

It is plausible that in its beta versions, language was stunningly unclear, and you would just utter single words like "bear", leaving it to the listener to figure out that, most likely, there is a bear approaching and it's time to run. I would not consider uttering the word "bear" to be a command, even if it resulted in a hearer performing an action (presumably one that I wanted him to perform). Maybe by "command" you mean an utterance where the speaker wishes the hearer to do something, regardless of linguistic form. But then "commands" vastly predate language or mammals, since many animals utter sounds (or smells or colors) in order to get another animal to do something.

A problem with equating morphological simplicity with phylogenetic utility is that it predicts that objects (accusatives) should be the unmarked case, but in fact, nominatives (agents) are more often the unmarked case.


How to explore such things if there are no data. The evolution of languages may cover a period of a hundred thousand years or more. A lot of books with a lot of theories have been written but all this is speculation. The beginnings of language are a great mystery.

Jean Auel in her series of The Clan of the Cave Bear has presented some useful ideas about the language of the Neanderthals. In the beginning utterances were connected with a lot of gesture signals. But the Neanderthals appear late in the evolution of language.

If we observe dogs we can hear and see if a little dog is wailing because it is left alone in front of the supermarket or if it is agressive. Without words the animal can convey information. Other animals like bees have developed a complicated system of signals to give information where good plants can be found.

Over thousands of years the beginnings of language were connected with sounds and signals. It must have been an important step when words and word classes were developed. I think at first there was no differenciation between noun and verb. You could connect a word with "I" then it was a verb or with "many/no" then it was a noun. It is interesting to think about the beginnings of language - but we know nothing.

Not the answer you're looking for? Browse other questions tagged or ask your own question.