How the Romans invented grammar

« previous post | next post »

It's not only in the United States that linguists have failed in their responsibility to educate the public. As Geoff Pullum explained yesterday, the English Teachers Association of Queensland (Australia) recently published a teachers' guide to grammar that was "full of utter howlers". And some of the discussion of the controversy is not much better. For example, Graham Young wrote today in a blog post at the National Forum ("Grammar's taught to grammarians", 6/14/2008):

The Romans, driven I suspect by their infatuation with standardisation (which palls in comparison to ours, but they caught the disease first), invented grammar. It didn't exist before them, people just spoke languages.

This short passage contains several implicit indictments of my profession's educational failures.

The first is that among the several meanings of the word grammar, Mr. Young fixes on one — the least important one — while ignoring the others. Here's the American Heritage Dictionary's entry for grammar:

1a. The study of how words and their component parts combine to form sentences. b. The study of structural relationships in language or in a language, sometimes including pronunciation, meaning, and linguistic history. 2a. The system of inflections, syntax, and word formation of a language. b. The system of rules implicit in a language, viewed as a mechanism for generating all sentences possible in that language. 3a. A normative or prescriptive set of rules setting forth the current standard of usage for pedagogical or reference purposes. b. Writing or speech judged with regard to such a set of rules. 4. A book containing the morphologic, syntactic, and semantic rules for a specific language. 5a. The basic principles of an area of knowledge: the grammar of music. b. A book dealing with such principles.

Mr. Young thinks that the Romans invented grammar 3, "A normative or prescriptive set of rules setting forth the current standard of usage for pedagogical or reference purposes". It's obvious that the Romans didn't invent grammar 2, "The system of inflections, syntax, and word formation of a language", since every variety of every language ever spoken has such a system. So the first symptom of my profession's failure is that Mr. Young thinks that the Queensland controversy is (only) about teaching usage standards, and not about teaching the concepts and skills needed to analyze English of all kinds — standard, archaic, vernacular, whatever.

But in fact, the Romans didn't invent grammar 3, and neither did they invent grammar 1, "The study of how words and their component parts combine to form sentences". The earliest systematic study of a linguistic system that has come down to us is that of Pāṇini (4th or 5th century BC), whose commentators reference several earlier accounts of Sanskrit grammar, such as Śākaṭāyana (8th century BC). The goal of this tradition was to preserve knowledge of the language of the Hindu religious canon, bccause in Panini's time, the language in everyday usage had changed so much (since the composition of works like the Vedas) that correct recitation and understanding of the sacred works could not be assured without explicit study.

There are some even earlier examples of grammar 3 and grammar 1: you can read Stephen Langdon's "Sumerian Gramatical Texts" online. These "grammatical texts belong chiefly to a large group of tablets known as school texts. They represent the pedagogical books and pupils’ exercises of a Sumerian college", from the Nippur Collection, dated to the period between 3200 BC and 1800 BC. These tablets were used to teach Sumerian vocabulary and morphological inflections to students whose native language was Akkadian.

When the Romans finally got around to grammar 1, in the first century B.C., what they did was almost entirely copied from earlier Greek models. Mr. Young attributes his historical information to Nick Ostler's excellent "Ad Infinitum: A Biography of Latin", so I'll quote some bits of Ostler's account, which may also help to explain how Mr. Young went astray:

The Greek language had progressively been analyzed since the fifth century BC, first by the sophistic rhetoricians and philosophers of Athens, who tended to look for general principles […], later by Stoic philosophers and Alexandrian textual critics, who emphasized more the arbitrary and irregular […]. The two aspects were characterized by the Greeks as "analogy" and "anomaly" and theorists disputed in vain which of the two was truly fundamental to language. […] [T]he traditions culminated in the first comprehensive textbook of Greek grammar, Dionysius the Thracian's Tekhne Grammatike 'The Scholarly Art', written around 100 BC. […] At the time, the aim of these studies was said to be the criticism of literature.

[…] The analysis of Latin on Greek principles took off with … M. Terentius Varro (116-27 BC), whose DE LINGVA LATINA 'On the Latin Language' included a treatment of Latin inflexion, and who can often be seen thinking like a modern formal linguist.

The analysis continued to be elaborated, and simplified versions came to be included in the grammatical syllabus. Q. Remnius Palaemon, a famous practitioner of the first century AD, incorporated most of Greek terminology into Latin in translated form, including the famous mistranslation of aitiatike, the 'caused' case, as ACCVSATIVVS 'accusative'.

[…] Scholars' adaptations of grammatical theory to Latin gave the language a new source of status, putting it effectively on a par with Greek even at this, most abstract level. But there was another motivation for developing grammar, one that brings us back to the schoolroom. Foreigners aspiring to learn the language well, especially as it began to change, needed instruction on what was good style; seeing examples of it held up for imitation was no longer enough for learners. Grammatical theory began to be presented, often in simplified form, in the classroom. The word barbaros / BARBARVS came to at least as commonly used to denigrate failures in grammar and style (in Greek or Latin) as to point something out as truly foreign. A. Gellius, a scholar of the second century AD, naturally described a correct usage as NON BARBARE DICERE, SED LATINE 'saying it not barbarously but in Latin'.

And while such implicit snobbery against the outsiders continued to prevail, a curious fact was missed. Already by the first century AD, Latin scholars had demonstrated that Greek was not the only language reducible to rule, even if those very rules were inspired by looking at Greek. Other languages too could have a grammar.

Yet it would be another millennium and a half before Europeans would realize the implications of this for languages at large.

Call it two millennia, and counting.


  1. john riemann soong said,

    June 14, 2008 @ 12:48 pm

    I see you also refrained from commenting on the ungrammaticality present in the sentence, " It didn't exist before them, people just spoke languages," what with two independent clauses and all.

    There's also some hint of Eurocentric bigotry in that comment that irks me and makes me want to pound the author's head with a hammer.

    But I am curious Mark — though I do agree with you that public awareness of linguistic concepts needs to be raised — how do you suggest we address this problem? Do we start teaching linguistic concepts to primary school children, or use IPA to teach phonemic awareness at that age? Reinvent how grammar is presented at the primary and secondary levels? Form a "Society for Public Linguistic Awareness" and run ads on television? ("Friends don't let friends be Vulgar Prescriptivists." / "Live above Vulgar Prescriptivism. Live Above the Influence.")

  2. Spectre-7 said,

    June 14, 2008 @ 2:03 pm

    John: Correct me if I'm wrong, but I don't think that's actually an error of grammar. Both clauses are well formed and entirely comprehensible. If spoken aloud, no native speaker would have any difficulty in interpreting their meaning. The issue instead seems to be one of orthography (punctuation specifically), in that modern style calls for either a period or a semi-colon to separate independent clauses.

    In any event, one of the things I appreciate most about the scholars who walk the hallowed halls of Language Log Plaza is precisely that restraint in pointing out minor flaws in writing. They much prefer (as far as I can tell) to point out major flaws in thinking. ;)

  3. rootlesscosmo said,

    June 14, 2008 @ 3:56 pm

    Also: "palls in comparison to ours"? I understand "pall" to mean "become stale, wearisome;" I would have used "pales" to mean (as Young seems to mean) "appears insignificant."

  4. john riemann soong said,

    June 14, 2008 @ 7:05 pm

    Spectre-7: Hmm, it's actually an issue I've been thinking about. Can someone else chime in on this? For example, is "your off the mark" a grammatical error, or an orthographic one?

    Also, since I don't have any background in English phonology, can someone tell me if it has been shown whether there are significant differences in the timing of pauses traditionally marked by commas and periods?

    Without a semicolon or a period, reading that sentence was "jarring" for me, mentally suggesting that there was an element of "ungrammaticality" to it. If someone said to me aloud, "It didn't exist before them people just spoke languages" with the pause between the two clauses being comparable the pauses between the rest of the words in that sentence, it would be jarring, I think. Suprasegmental features of phonology — like pauses — are part of grammar, aren't they? It seems to me that there are constraints that dictate when one must make them and when one should refrain from making them.

    Of course this is a rather minor concern, compared to yes, Young's rather major error in thought. :-)

    "Also: "palls in comparison to ours"? I understand "pall" to mean "become stale, wearisome;" I would have used "pales" to mean (as Young seems to mean) "appears insignificant."

    In Mr. Young's accent, would pall/pale be minimal pairs? Australian English isn't uniform, is it?

  5. Mark Liberman said,

    June 14, 2008 @ 7:57 pm

    The sentence "It didn't exist before them, people just spoke languages" is indeed an example of a comma splice, sometimes also called a "comma fault". I'll limit myself to quoting Lynne Truss on this one: "so many highly respected writers observe the splice comma that a rather unfair rule emerges on this one: only do it if you're famous." (By "observe" she seems to mean "use".)

    And "palls in comparison to" is indeed an example of the spelling-confusion type of malapropism — the idiom that Young was reaching for was certainly "pales in comparison to".

    Young positions himself against prescriptivism, as well as against the idea that anyone ought to be taught grammatical analysis. He combines the two arguments in a conceptually confused and historically inaccurate way, as I tried to explain. In that context, it's a distraction to discuss his infelicities of expression.

  6. marie-lucie said,

    June 14, 2008 @ 10:29 pm

    As a student of English as a second language, I was never warned about the comma splice, and I never read a definition of it, so that for a long time I was puzzled about what it was that English professors were getting so het up about (it seemed to be a tossup between the comma splice and the split infinitive). As time went on I formed the impression that the "comma splice" referred to cases where the word that omitted, as in "it was so hot, we just had to jump into the pool". I must have been wrong, since in such a case, adding "and" or replacing the comma with a colon are not possible. But I see from the example at the end of the quoted paragraph that I too must have been happily committing comma splicing for years, without anyone calling my attention to this allegedly grievous sin.

  7. john riemann soong said,

    June 14, 2008 @ 11:20 pm

    I note from the Wikipedia article:

    "Strunk & White note that splices are sometimes acceptable when the clauses are short and alike in form, such as:

    The gate swung apart, the bridge fell, the portcullis was drawn up."

    Disregarding the fact that this observation comes from a prescriptivist style manual, it appears to me that this sort of comma-spliced sentence "reads naturally" because the short length of each clause means there isn't much pause between each clause anyway. If punctuation is meant to represent prosody, I would say that the puncutation present in

    "It didn't exist before them, people just spoke languages"

    represents a prosody that seems to me rather non-standard.

    Does non-standard prosody imply ungrammaticality? It's a tangential question, but one that makes me curious.

  8. Phil said,

    June 15, 2008 @ 1:37 am

    It is alarmingly widespread, this idea that you have to teach language to children, and that the way to teach them is through prescriptive admonishments, particularly by "correcting" children's "mistakes". My brother is a lecturer in education and he has been heard to espouse this view (he specializes in religious education, but still…)

    The problem seems to lie in a failure to fully appreciate the distinction between language acquisition and language teaching; perhaps that's the big concept that Mark and colleagues should be stressing to the media whenever they get the chance.

    If you understand acquisition, then it would surely help you to see clearly what prescriptive rules should be (basically, style suggestions rather than statements about language), and to understand that language is more usefully viewed as a natural phenomenon than it is as an artifact.

  9. Christopher Stone said,

    June 15, 2008 @ 3:40 am

    "But I am curious Mark — though I do agree with you that public awareness of linguistic concepts needs to be raised — how do you suggest we address this problem?"

    Obviously I can't answer for Mark, but at the university I currently attend (Truman State University in Kirksville, MO), there's a summer nerd camp called Joseph Baldwin Academy where academically advanced students in the 7th, 8th, and 9th grades come for three weeks and get the equivalent of a semester course in whatever class they're taking. Last year we had a course called "Language Myths and Truths", which was essentially Intro to Linguistics plus about half a semester of generic Sociolinguistics. It was an amazing experience for all involved, especially the AAVE speaker and kid from Alabama that we had – they both got some nice linguistic tools to have some good self-esteem about their dialects while also getting the tools to analyze other people's linguistic arguments.

    The class itself only had about 12 kids, so obviously it isn't going to change the course of linguistic prejudice in the US, or even in their schools, but programs like that one are definitely tools for spreading the gospel of linguistic science (juxtaposition intended).

  10. Mark S. said,

    June 15, 2008 @ 5:28 am

    The Romans invented grammar? Ah, that might explain why "Chinese grammar is an import from the West."

    That's from a regular writer for one of Taiwan's three English newspapers. The rest of the piece is filled with similar nonsense.

  11. Leon said,

    June 15, 2008 @ 8:31 am

    If you understand acquisition, then it would surely help you to see clearly what prescriptive rules should be (basically, style suggestions rather than statements about language), and to understand that language is more usefully viewed as a natural phenomenon than it is as an artifact.

    Spoken language may be more "natural" than "artificial" (usually), but written language is a different matter. Although I'm writing this sentence based on an intuitive understanding of the grammatical rules governing (fairly formal) written text, if I was you know writing like I speak, then ma- maybe I wouldn't be so clear or easy to understand if you see what I mean or maybe the stylistic differences would be jarring mmm?

    So sure, there's a sense in which I "acquired" my ability to write formally, but it was by reading lots of formal English texts, writing some for myself, and — usually in a classroom context — being "prescriptively admonished" by teachers if that writing didn't conform to formal standards. Osmosis and a love of learning played their part, but I was also taught. And without this ability, I wouldn't be able to participate as fully in certain parts of society, such as political debate, academia, and blogs like this one. Call them "stylistic differences" if you'd like; either way, the "style" of standard/formal/good/clear English is important for a well educated person to understand.

    P.S. I tend to use a semicolon rather than a comma splice; though I picked up the habit by osmosis, I've confirmed it by finding examples in well-edited writing.

  12. David Marjanović said,

    June 15, 2008 @ 9:10 am

    I'd probably prefer using a dash instead of a comma in the example, but apart from that I can't find anything strange about it. The pitch goes up during "before them", and then a pause follows in slow speech.

    Speaking of commas, "I am curious Mark" means that your name is "curious Mark".

  13. David Marjanović said,

    June 15, 2008 @ 9:17 am

    For example, is "your off the mark" a grammatical error, or an orthographic one?

    Given the fact that nobody ever writes *"my off the mark" or *"his off the mark", I vote for an orthographic one.

  14. Phil said,

    June 15, 2008 @ 11:57 pm

    Well, you're most of the way to understanding the distinction.
    You say that style and register are important, and I think everyone agrees with this. You also say that you were (in part) *taught* your formal English writing style. Here we have to be very careful about what exactly you were taught. The process of learning a formal style was not a process of learning language. It was a process of selecting which types of language (vocabulary, sentence structure, discourse structure) to apply in a situation. Your total knowledge of English vocabulary and structures was (to a very great extent) unchanged at the end of the process of learning formal style. What you had gained was a greater appreciation of which bits of the language you knew were appropriate in one particular social setting (academic writing).

    That kind of learning is in contrast to what my one-year-old is doing right now. Yesterday, for the first time in my hearing, he spontaneously put together an utterance with two parts: baba, zei (Dad, sit down – he was speaking Minnanhua). This represents an acquisition of a new language ability.

    Both of these two kinds of learning can occur without deliberate teaching – in which case it's usually called acquisition. And the learning of a style can probably be accelerated by good teaching. It's open whether acquisition of new language can be helped by deliberate teaching. From what I've read, I suspect not, with the possible exception of some kinds of vocabulary. But it's certainly true that most of the acquisition of language by most children in the world occurs without deliberate teaching.

    Also important is that your formal English style uses a large subset of the same grammar that you use in your speech. There's no written grammar that you learned for writing because you didn't have it in your spoken English grammar. Spoken language, like any natural phenomenon, can be analysed (imperfectly), and the resulting conclusions are called (descriptive) grammar. Written grammar is just a set of rules about which bits of our spoken grammar should not be used in written language.

  15. Tearlach said,

    June 16, 2008 @ 3:58 am

    If we are so concerned about "educating" the public about linguisticism, then we should also make a distinction between linguistics and grammarians. The difference equals the difference between an economist and an accountant. One asks important questions about the scientific nature of a thing, and the latter counts beans.

  16. Elizabeth Moon said,

    June 16, 2008 @ 12:29 pm

    In answer to John Riemann Soong's question about length of pauses signaled by commas and semicolons:

    From the POV of the professional writer and the individual who records written work for audio books and the like, there is indeed a difference in the pause length between words without punctuation, words/phrases separated by commas, by semi-colons, by colons, by dashes or ellipses, and by periods. These different punctuation marks also affect emphasis within preceding and following groups of words. Punctuation replaces the auditory signals (including, but not limited to, pauses of varying lengths) of ordinary verbal communication with visual codes.

    On language learning v. teaching: The neurologically typical child learns spoken language as part of normal maturation, but the use of the appropriate writing modalities requires some intervention. Punctuation marks, for instance, do not call attention to the reasons for their use. Even a child who reads a lot and writes a lot will not, on his/her own, acquire full competency in the use of punctuation.

    Neurologically atypical children are a special challenge…my one-data-point observation of our son (autistic, now 24 yo) leads to the conclusion that–to the extent the child's sensory apparatus allows–the sequence of language learning is similar but not identical to that of neuro-typical children. (I do wish we'd had the money for serial PET scans to see if what seemed to be a switch from right-brain auditory processing and so-called "right brain language"–a term one of his therapists used–to left-brain/Broca's area processing and complex syntax was reflected in that way. The social effect was that his burgeoning savant musical talent vanished as the language use soared. A good trade, in terms of long-term goals like independent living, but I miss the astonishing little pianist.) Interventions based on the concepts of linguistics were much more useful than ordinary speech therapy.

  17. John Greenfield said,

    June 17, 2008 @ 3:05 am

    The Romans "invented" grammar? Well that is one more to add to their other "invention"; cement. :) But seriously, where do these people get this tosh? This news would have shocked the young men skulking around the agora and gymnasium of fifth century Greece. What might we call the instruction these young men received – from the likes of Gorgias, Socrates, and Protagoras – in grammar, rhetoric, music, geometry, epic and lyric poetry, and the finer arts of 'giving head'? And what of all those Socratic dialogues finessing correct lexis And what of Plato's Academy, Aristotle's Lyceum, Gorgias, Protagoras, etc?

    I have just finished as essay on fifth century tragedy. I got very excited by Euripides masterful use of epideictic rhetoric. This excitement was largely generated by Aristotle's discussion of the role of diction – which includes GRAMMAR – in Poetics

    Anybody who thinks the Romans invented grammar has obviously never stumbled upon the delights of weak-aorist or semi-deponent verbs in Homer or Herodotus' deliberate use of intrusive oblique infinitives to augment his proleptic narrative strategies. Oh, and by the way, even the Romans knew it was Aristotle who was responsible the categorisation of "oblique" verbs, which are common to all Indo-European languages – Greek, Latin, German, Celtic, Sanskrit, slavic, Iranian.

    This bizarre Leftist fantasy that teaching grammar in schools is somehow another step in a vast right-wing conspiracy speaks volumes about their own education.

  18. Andy Hollandbeck said,

    June 17, 2008 @ 2:57 pm

    Tearlach: "We should also make a distinction between linguistics and grammarians. The difference equals the difference between an economist and an accountant."

    When tax time rolls around, which would you rather have: an economist or an accountant?

    If you want to become a good speaker or writer, who do you want: a grammarian or a linguist?

    I think the problem is that we, as a whole, don't really know what "grammar" means. We know that it's something smaller than linguistics, but something larger than just word order. But that leaves a whole, big, gray area.

    On top of that, there's no such thing as a universally accepted authority on grammar — at least not in English, and I suspect in every other language but French as well.

  19. Tearlach said,

    June 18, 2008 @ 7:16 am

    I don't equate grammar with good speaking or writing any more than I equate orthagraphy with intelligence.

    And to John Greenfield, I myself don't equate grammar obsession with the vast right wing conspiracy. Because a) I associate the current vast right wing conspiracy with anti-intellectual populism, not old school elitism. And b) I DO associate the grammarly obsessed with the young very feminist and oh-so PC ladies of the English department. That's just my anecdotal evidence, nothing more.

RSS feed for comments on this post