Grammar in schools
« previous post | next post »
"Rules for teaching grammar in schools", The Economist 3/12/2022 ("It may not make children better writers. But it is valuable all the same"):
Absence of evidence is not, as the saying goes, the same thing as evidence of absence. But if you continue looking for something intently, and keep failing to find it, you can be forgiven for starting to worry. And so it is with the vexed—and in Britain, highly politicised—subject of explicit grammar teaching in schools, and its link or otherwise with improved writing ability.
Another study, in this case a large randomised controlled trial, has recently been added to the expansive literature on the subject. Like nearly all its predecessors, it found that teaching kids how to label the bits and pieces in a sentence does not make them better writers.
[…]
In retrospect it scarcely seems surprising that learning to underline a modal verb, such as “can”, “should” and “may”, does little to help students use them effectively in their own writing. These words are anyway grasped by tiny children without the need to know what they are called. This may tempt the conclusion that the teaching of grammar should be shelved altogether. But there are reasons to reform it rather than scrap it.
Understanding of language is part of a wider education in what makes human beings human.
Back in 2013 ("School grammar, round two"), I wrote:
Quite a few commenters asked whether grammar instruction would "be sufficient to produce high school graduates capable of composing a cogent paragraph made up of more or less grammatical sentences". The relevance of grammar instruction to second-language teaching and learning also came up. These are interesting and relevant questions, but I feel that it would be a mistake to make them the center of the discussion. We don't put chemistry into the school curriculum because it will make students better cooks, or even because it might make them better doctors, much less because we need a relatively small number of professional chemists. We believe (I hope) that a basic understanding of atoms and molecules is knowledge that every citizen of the modern world should have.
I feel that the arguments for grammar in the school curriculum — and for linguistic analysis more broadly — ought to be similar. A basic understanding of how language works should be part of what every educated person knows. And there are many professions where a more-than-basic understanding is worthwhile, just as pharmacists and farmers need quite a bit of advanced practical chemistry to do their jobs well.
The cited Economist article puts it this way:
There are practical reasons to ask children to grapple with grammar, too. One is that an explicit knowledge of it will make learning a foreign language easier. Even if you did intuit how to make subordinate clauses in your native languages as a toddler—just without instruction—getting to grips with them in German or Russian in later years is simpler if you know how to define and spot them. As it is, many English-speakers come to understand grammar by studying a foreign language, rather than the other way round.
For grammarians keen on the jobs of the future, the field of natural-language processing is booming. After many years of poor results, technological wizards have devised programs for automated translation, speech recognition (as in dictation software) and other services that are actually usable, if far from perfect. These tools may rely more on knowledge of artificial intelligence than of the subjunctive, but linguistic expertise still matters, and may give budding programmers an edge over rivals whose best language is Python.
I'd also add that a practical understanding of (an appropriate approach to) discourse analysis — what used to be called "rhetoric" — would probably be more effective than sentence-diagramming as an aid to (at least some) apprentice writers. And lawyers should probably be higher on the list of relevant professions than programmers…
Update — ktschwartz in the comments reminds us that it's always fatal to believe mass-media reports of (allegedly) scientific papers, without careful scrutiny of the original publication(s).
According to a blog post by Stroppy Editor:
A recent study, as you may or may not have heard, has found that teaching grammar to Year 2 children (age 6-7) does not improve their writing. But that’s not what it found.
The study, by researchers at UCL and the University of York, did not compare grammar teaching with no grammar teaching. It compared one particular programme of grammar teaching, called Englicious, with the grammar teaching that schools were doing already, and it found that Englicious produced results that were essentially no better than the other grammar teaching.
The research paper makes this very clear, although the conclusions section seems to stray a little into over-generalisation.
As the cited blog post points out, the egregious over-interpretation apparently started (as often) with the press releases from the two universities involved. This unholy alliance of university PR people and journalists does a lot of damage. I should have known better than to trust them — I relied on the Economist's "Johnson", who is usually a reliable source, but in this case seems to have been led down the garden path by the press releases and other journalists…
Some maybe-relevant past LLOG posts:
"The plastic fetters of grammar", 10/21/2003
"Personal and intellectual history of sentence diagrams", 10/14/2004
"Grammar: Carrot or stick?", 7/4/2008
"Grammar school", 4/14/2008
"A grammar book in grammar school?", 2/18/2009
"Law as applied linguistics", 7/25/2009
"Grammar, time and truth", 4/18/2011
"Corpus linguistics in a legal opinion", 7/20/2011
"Nominee for the Trent Reznor Prize", 4/14/2012
"Diagramming Sentences", 4/14/2013
"Misreading like a lawyer", 9/27/2013
"Putting grammar back in grammar schools: A modest proposal", 12/25/2013
"School grammar, round two", 12/30/2013
"Sentence diagramming", 1/1/2014
"William Hazlitt on grammar", 12/17/2014
"Three cheers for Michael Gove", 12/28/2014
"Grammatical politics", 12/31/2015
"'But I was going to say that but now I won't say it'", 7/30/2016
"Lawyers should learn linguistics, part infinity", 12/13/2016
"Lawyers as linguists", 10/21/2020
Benjamin Orsatti said,
March 14, 2022 @ 7:51 am
Prof. Liberman,
Regarding the following:
"'[L]inguistic expertise still matters, and may give budding programmers an edge over rivals whose best language is Python.'" [Quote from Economist]
"[A] practical understanding of (an appropriate approach to) discourse analysis — what used to be called "rhetoric" — would probably be more effective than sentence-diagramming as an aid to (at least some) apprentice writers. And lawyers should probably be higher on the list of relevant professions than programmers" [ML]
Actually, the occasional comma-decided-case notwithstanding, grammar, insofar as it serves as an aid to statutory construction or contract interpretation, doesn't really play as big a role as you might think in law. Much more frequently employed are the various common law and state law statutory "canons of statutory construction" (i.e. for statutes) and treatises/hornbooks, like Elkouri & Elkouri's "How Arbitration Works" (i.e. for interpretation of labor contracts).
On the other hand, if a vote were held today as to whether rhetoric should be a mandatory first-year law school course, I'd readily register in the affirmative.
[(myl) I agree that comma-placement is not a great reason for lawyers to understand linguistic analysis. But interpreting linguistic meaning in context is surely a central issue, involved in the evolution of "common law" as well as in ordinary life. And aside from the ubiquitous question of "what word X means" in general, there are issues like the old distinction between de re and de dicto interpretations, which even has Latin terminology to recommend it! (See "Misreading like a lawyer" for a practical example.) My outsider's impression is that there are dozens of issues like this, where a small amount of analytic sophistication would stand many lawyers in good stead. ]
J.W. Brewer said,
March 14, 2022 @ 7:58 am
Learning useful things about your L1 grammar (not always in the best conceptual schema) as a side effect of formal school instruction in an L2 is a very traditional approach, going all the way back what were already called in medieval times "grammar schools" in England, meaning schools where you explicitly learned Latin grammar. One relevant factor to how well this approach might work, of course, is the age at which children in Anglophone schools will begin formal instruction in an L2, which the Economist piece implies is likely to be "in later years" compared to whatever age they think explicit description of suburdinate clauses in English should be taught.
Cervantes said,
March 14, 2022 @ 9:52 am
Learning Spanish certainly gave me insight into grammar, but a lot of it consisted of discovering what English does not have. As English is a creole, its grammar is degenerate. I don't mean that in a bad way, it's actually interesting that it works semantically without much of the syntactical machinery present in other languages, and which English originally possessed.
Benjamin Orsatti said,
March 14, 2022 @ 11:23 am
Prof. Mark Liberman said:
"[I]nterpreting linguistic meaning in context is surely a central issue, involved in the evolution of "common law" as well as in ordinary life. […] My outsider's impression is that there are dozens of issues like this, where a small amount of analytic sophistication would stand many lawyers in good stead."
—
Well, there are at least 19 in Pennsylvania, in any event:
1 Pa.C.S.A. PA ST Pt. V, Ch. 19, Subch. B. Construction of Statutes
§ 1921. Legislative Intent Controls
§ 1922. Presumptions in Ascertaining Legislative Intent
§ 1923. Grammar and Punctuation of Statutes
§ 1924. Construction of Titles, Preambles, Provisos, Exceptions and Headings
§ 1925. Constitutional Construction of Statutes
§ 1926. Presumption Against Retroactive Effect
§ 1927. Construction of Uniform Laws
§ 1928. Rule of Strict and Liberal Construction
§ 1929. Penalties NO Bar to Civil Remedies
§ 1930. Penalties for Each Offense
§ 1931. Intent to Defraud
§ 1932. Statutes in Pari Materia
§ 1933. Particular Controls General
§ 1934. Irreconcilable Clauses in the Same Statute
§ 1935. Irreconcilable Statutes Passed by Same General Assembly
§ 1936. Irreconcilable Statutes Passed by Different General Assemblies
§ 1937. References to Statutes and Regulations
§ 1938. References to Public Bodies and Public Officers
§ 1939. Use of Comments and Reports
Speaking of grammar specifically:
§ 1923. Grammar and punctuation of statutes
(a) Grammatical errors shall not vitiate a statute. A transposition of words and clauses may be resorted to where a sentence is without meaning as it stands.
(b) In no case shall the punctuation of a statute control or affect the intention of the General Assembly in the enactment thereof but punctuation may be used to aid in the construction thereof if the statute was finally enacted after December 31, 1964.
(c) Words and phrases which may be necessary to the proper interpretation of a statute and which do not conflict with its obvious purpose and intent, nor in any way affect its scope and operation, may be added in the construction thereof.
But, to your original point, yes, lawyers could _always_ use more "analytic sophistication" and less "synthetic sophistry".
J.W. Brewer said,
March 14, 2022 @ 11:38 am
As part of the very small minority of working American lawyers with an academic background in linguistics I think the most practical positive impact of more lawyers having at least a good college-intro-course might be a reduction in the percentage of people in the profession with stupid prescriptivist peeves and associated hang-ups who cause lawyers time and clients money being wasted by editing documents to avoid split infinitives or obey zombie rules about that v. which, etc etc.
Problems that arise from badly-drafted contractual or statutory language tend IMHO to often arise from factors that academic linguistics in its current form is not necessarily well-positioned to solve, i.e. some mix of: (a) multiple drafters/editors of a single document creating a too-many-cooks-in-the-kitchen dynamic instead of a single speaker/writer of the language with an actual subjective idea in their poor monkey brain that they were trying to express themselves clearly; (b) the need to write about things of sufficient complexity that our poor monkey brains are just not good at putting them into natural language; (c) long and complex documents being finalized (sometimes after a lengthy prior drafting process, sometimes not) under considerable time pressure, which predictably leads to failures in proofreading and quality control especially because Someone Important will often have made significant last-minute changes in one place whose potential ramifications elsewhere are not immediately obvious; and (d) having language that seems clear enough in isolation lead to confusion because it interacts badly, in the context of a long and complex document, with other language that seems clear enough in isolation, where there is again the problem that no single human's poor monkey brain can simultaneously think about all the different moving pieces in what the document is trying to do — i.e. you deal with (b) by breaking the overall document down into manageable pieces but making sure all of those individual pieces interact appropriately at the macro level requires a skill set that does not come out of ordinary daily fluency in a natural language,
/df said,
March 14, 2022 @ 11:39 am
The London Sunday Times ran a piece by a teacher turned author (her memoirs attracted weird taunts of racism for mentioning one former pupil's "almond-shaped eyes" although the pupil herself was more than fine with it), that favoured teaching children to write by osmosis from set texts to the exclusion of explicit grammar lessons.
The article was supplemented by the paper's canned guide to grammatical concepts, all familiar to those who learnt grammar in the Latin lesson, apart from the notorious "fronted adverbial". In the spirit of Monty Python's Serbo-Croat phrasebook, it gave this example of a "subordinate clause":
"Writing, they knew, was a deep, mysterious craft, **and not for the likes of them**."
Possibly this writer also can't identify a passive verb (as regularly discussed here); or if grammar must be learnt from reading, can anyone formulate his or her own definition of a "subordinate clause", including one that is not a clause at all?
/df said,
March 14, 2022 @ 11:41 am
Obvs s/Serbo-Croat/Hungarian/
Stephen Cash said,
March 14, 2022 @ 11:46 am
As a lay-person and linguistic enthusiast, my completely non-expert impression is that a great number of people who have not had formal linguistic instruction, when asked if they think it might be valuable (for them or their children), feel like it would not be of value because they "already know how to speak English." Rumsfeldianly, they don't know what they don't know. And at least in my experience declarations of "there's a lot more going on than you realize" are met with, shall we say, skepticism.
Peter Grubtal said,
March 14, 2022 @ 12:52 pm
In 12 years of school education the time needed to be spent on the basics of grammar (parts of speech, sentence structure) should scarcely figure as an enormous burden on the curriculum. And the children who are going to start a foreign language, hopefully a large proportion, are going to have to do it anyway. It seems that being opposed to the teaching of grammar in English schools is a shibboleth (and has been for some time), a sign of one's progressiveness or political persuasion.
Matthew L Juge said,
March 14, 2022 @ 1:31 pm
@Cervantes The English/creole hypothesis has not attained consensus status.
ktschwarz said,
March 14, 2022 @ 2:08 pm
Unlike all the journalists, Stroppy Editor (Tom Freeman) actually read the research paper, and points out something obvious that the coverage missed or downplayed:
(Boldface added.) He also observes that the study only compared "test scores from before and after a ten-week period of teaching" — ten weeks isn't much!
Philip Anderson said,
March 14, 2022 @ 3:08 pm
@J.W. Brewer
“Learning useful things about your L1 grammar (not always in the best conceptual schema) as a side effect of formal school instruction in an L2 is a very traditional approach”
And perhaps the best approach.
Grammar is useful when learning an L2, but much less so when learning an L1. It’s much easier to understand grammar when comparing languages (a geography syllabus that only looked at the pupils’ home area would be pretty pointless). Learning that different languages use different constructions to express the same idea is vital to avoid the disaster of word-for-word translation. There may some use for grammar if there’s a big difference between a home dialect and the standard language.
In the UK formal language instruction tends to start late, unfortunately, and certainly after the grammar years.
J.W. Brewer said,
March 14, 2022 @ 4:49 pm
Another key point in Tom Freeman's discussion that ktschwarz linked to is that the only-ten-weeks-long trial comparison was done on students in "Year Two" which I believe is BrEng for "first grade." If I were going to pick a population of students to see whether the quality (or even grammaticality) of their writing increased after exposure to such-and-such proposed new approach for explicitly teaching grammar, I'm not sure that's the age group I'd start with.
Perhaps it made sense in the UK context to start there because of an existing top-down mandate to have *some* sort of explicit grammar instruction at that age, meaning you naturally want to figure out the least-bad way to do it, but I suspect you might well get more bang for your buck (your quid?), or at least see a wider range of outcomes between different pedagogical approaches, at a somewhat older age.
Viseguy said,
March 14, 2022 @ 7:00 pm
@J.W. Brewer: With your points (a) through (d) above, you've got the makings of a valuable elective writing course for law school — or, better still, a unit within the required writing course. My first-year writing course, taught by a tax lawyer, was transformative, but it didn't prepare me for the reality I encountered on joining a law firm: "Abandon pride of authorship, all ye who enter here." Although year after year of having my "work product" torn into and thrown back at me made me a better lawyer and (I think) a better writer, it did not alert me to the pitfalls of writing-by-committee which you aptly point out. Which leads me to ask, Is Law and Linguistics a thing these days? I took two semesters of Law and Psychology (nearly 50 years ago), but your post leads me to believe that L&L would have been equally valuable.
Stephen L said,
March 15, 2022 @ 1:44 am
>Learning useful things about your L1 grammar (not always in the best conceptual schema) as a side effect of formal school instruction in an L2 is a very traditional approach
I wonder if this generally leads people to have a feeling of their mother tongue being impoverished because they're not aware of the specificities of it.
> There may some use for grammar if there’s a big difference between a home dialect and the standard language.
I've seen a video clips of english high-school classes that talk about english dialects and the discussions were both quite intricate on topics of syntax, and also because people love throwing in their two-pence about L1 issues (and can work as oracles for it), the kids were able to chat really productively about it – maybe this is the best 'in' to teaching 'L1' grammar in schools? (Rather than just teaching grammar of a standardised version of the language)
Philip Taylor said,
March 15, 2022 @ 4:36 am
Philip A — "In the UK formal language instruction tends to start late, unfortunately, and certainly after the grammar years" — In that case, things have changed fairly dramatically (and for the worse, INHO) since I was at school. In the last couple of years at primary school (so between the ages of 9 and 11) we learned the basic parts of speech (nouns, verbs, adjectives, adverbs, pronouns) and the basic structure of a sentence. At grammar school we learned to parse sentences, identify clauses and phrases, etc., as part of our tuition in English Language, the part of the English curriculum that I really enjoyed — the other part, English Literature, I loathed and detested. Needless to say, in my GCEs I passed English Language with flying colours and failed English Literature. Even today, at the age of 75, I still have no real interest in English Literature — Jane Austen, for example, remains a closed book to me — although I certainly enjoy Charles Dickens, JRR Tolkien and John le Carré.
Philip Anderson said,
March 15, 2022 @ 8:33 am
Philip T:
I was thinking more of foreign language teaching, although that is better than when I was at school.
I can only remember English grammar in secondary school, but my son is learning some in Year 2 (6-7) now. Useful but there are bigger influences on his writing at that age – reading, encouragement to use more varied language and descriptive words.
CD said,
March 15, 2022 @ 6:14 pm
Yes, there is something to be said for being able to pick out the verb in a sentence, and a few more basics. But I suspect the returns to instruction in L1 formal grammar fall off steeply after that.
Re "a basic understanding of how language works should be part of what every educated person knows," maybe, but (a) that's the fallback argument for retaining instruction in *any* damn thing (b) does the standard English-class sentence-diagramming and exhaustive inventory of speech parts really provide such an understanding? Whatever grip on the workings of language that I got as a kid was from L2s.
SusanC said,
March 21, 2022 @ 7:40 pm
As I understand it, there's a theory that we learn grammar rules by reading/hearing examples of them in use, rather than memorising an explicit formal definition of the grammar. (This is pretty obviously the case for children learn8ngb5heirvfirst language),
Thus, if you've been using sentences containing verbs, you already know how to use verbs, even if you have no idea that grammarians call them "verbs".
On this account of language acquisition, being taught e.g. that verbs are called "verbs", is not all that helpful. What you want instead is to read texts that exhibit the various grammatical constructs,