Also, check the back seat

« previous post | next post »

Reader FW asked for some advice about a nanowrimo discussion of "Ands and buts", which started Nov. 3 with this question:

So this is one that always get [sic] me.

Grammatically speaking, or however it is known, can you use Ands and Buts at the beginning of sentences? And can you use it at the start of dialogue as well?

A participant using the name pointytilly links to Paul Brians' list of "non-errors" in defense of the view that sentence-initial and and but are "grammatically correct". And indeed they are, according to essentially everyone with any plausible claim to expertise, prescriptivists and descriptivists alike.

As Arnold Zwicky wrote ("However,…" 11/1/2006):

Mark notes that the AHD [American Heritage Dictionary] note for and rejects NIC ["No Initial Coordinators"] out of hand, and he provides a smorgasbord of cites (and statistics) from reputable authors.  Similarly MWDEU [Merriam Webster Dictionary of English Usage].  Paul Brians, collector of common errors in English, labels sentence-initial coordinators a "non-error".  Bryan Garner denies, all over the place, that NIC has any validity.  Even the curmudgeonly Robert Hartwell Fiske tells his readers that there's absolutely nothing wrong with sentence-initial coordinators.  A point of usage and style on which Liberman and I and the AHD and the MWDEU stand together with Brians and Garner and Fiske (and dozens of other advice writers) is, truly, not a disputed point.  NIC is crap.

But still it lives on, as what I've called a zombie rule.  It's been lurking in the grammatical shadows for some time — at least a hundred years, to judge from MWDEU.  Hardly any usage manual subscribes to it, but it is, apparently, widely taught in schools, at least in the U.S., with the result that educated people tend to be nagged by a feeling that there is something bad about sentence-initial and (and but and so).  (It might well be that this sense of unease rises with level of education.  Someone should look at this possibility.)

Back on nanowrimo, Tom L Waters contributed this interesting counter-argument:

Allow me a quibble, pointytilly. The site you link to (which I am a fan of, by the way) is concerned with usage standards, not grammar. They are related but not identical. Grammar deals with categories such as parts of speech, and the logical rules of syntax for constructing sentences. Grammatically, conjunctions link words, phrases, or clauses. So from a grammatical standpoint, a sentence beginning with a conjunction is a fragment, and hence ungrammatical.

For many decades, grammar per se has become less and less relevant. From a linguistic standpoint, the rules of grammar are not adequate to the task of describing how language is actually used. Talking about acceptable usage rather than grammatical correctness avoids some of the problems created by grammar's emphasis on categories and rules, and recognizes the subtleties and complexities of natural languages.

Nevertheless, the rules of grammar remain one way that readers will evaluate the "correctness" of a piece of writing, so it's helpful to understand when grammatical rules are being broken, even if the usage in question is otherwise acceptable.

But if the "grammatical" definition of conjunction included sentences among the things to be linked, then sentence-initial conjunctions would be "grammatical", right? Well, the relevant definition in the OED is: "One of the Parts of Speech; an uninflected word used to connect clauses or sentences, or to co-ordinate words in the same clause." Merriam-Webster's definition is "an uninflected linguistic form that joins together sentences, clauses, phrases, or words".

Mr. Waters adds:

Although beginning a sentence with a conjunction is acceptable in fiction (there is wide agreement on this), it is not acceptable in technical, academic, or formal writing.

In dealing with someone who exhibits this level of zombie-like persistence, despite lists of devastating counter-examples from the most authoritative sources of formal writing, in the face of rational counter-arguments from every available grammatical authority, it's clear that the expertise that FW needs is not linguistic.  She should turn instead to Columbus's list of rules for surviving in a zombie-infested world, and perhaps especially #22, "When in doubt, know your way out".

And a little sunscreen never hurt anybody.


  1. T. A. Smith said,

    November 7, 2009 @ 10:44 am

    To prove your point: For about seventy years, I was painfully uneasy about writing a sentence that began with "And" or "But" or that ended with a preposition. The result: (estimate) 17,050 sentences that made sense but were embarrassingly awkward. But, being a slow learner (better than being a non-learner, I suppose), it was not until I reached seventy-five that I overcame my unease and practiced what I had long realized was a convenient and satisfactory usage.

  2. Liz said,

    November 7, 2009 @ 11:26 am

    Well, yes. But it does depend on whether you are teaching zombie rules because you believe they are true and unbreakable or because you are desperate to help people who have never been taught anything much that makes sense to write a letter that might actually help them get a job. I teach adult drop-outs from the education system who have been marginalised for years. Concepts like clause, phrase and even sentence are completely alien, and commas are favoured far more than full stops, particularly between subject and verb. The majority of employers are not sophisticated linguists, and may not be too understanding of a sentence like:

    I, use to work at X. But the boss was unreasonable. And I left. And I couldn't get another job.

    These students are not dim. They are just completely unable to cope with written English. They were probably told at some stage not to start sentences with And, and they see it used all the time, and do not have the skills to figure out when it works and when it doesn't.

    (I'm not a sophisticated linguist either. Would that last sentence be better with a full stop before each and? On what grounds?)

  3. Rachel Cotterill said,

    November 7, 2009 @ 11:35 am

    We were taught this at school; I think everyone is! Along with how to use an apostrophe, that's pretty much the only 'grammar' lesson that I can remember from compulsory education… but I ignore it.

  4. Karen said,

    November 7, 2009 @ 11:41 am

    "They were probably told at some stage not to start sentences with And, and they see it used all the time, and do not have the skills to figure out when it works and when it doesn't. (I'm not a sophisticated linguist either. Would that last sentence be better with a full stop before each and? On what grounds?)"

    Better how? I certainly wouldn't write it with a full stop before the "and", especially if that's the only change you're contemplating.

    I'd replace your first one with a "yet", myself, and precede it with a semicolon, and remove the comma before the second one. ("They were probably told at some stage not to start sentences with And; yet they see it used all the time and do not have the skills…") But that's just me.

  5. Liz said,

    November 7, 2009 @ 11:49 am

    Oh, and the people who write the grammar books are often the same people, or from the same camp, as those who set or mark test papers.

  6. TB said,

    November 7, 2009 @ 11:49 am

    I love these self-refuting rules. I suppose Mr. Waters doesn't realize that "although" is a conjunction.

    Also nice to see that the "students need our insane rules" argument so early in the thread.

  7. TB said,

    November 7, 2009 @ 11:52 am

    Or should I say "can be a conjunction"? I have a feeling I'm about to get schooled.

  8. Zoe Larivelt said,

    November 7, 2009 @ 12:18 pm

    Obviously it's a question of orthography, not grammar at all, since the supposed mistake can usually be fixed with a little repunctuation, without changing a word. "Grammatically incorrect" is just one of those things one says when something seems wrong, but it's too much trouble to figure out what.

    I wonder if the original proscription was simply against sentence fragments being allowed to stand alone. A silly rule, if you like, but at least it has a certain logic. And the zombie rule developed out of a kind of transference–since an initial conjunction frequently signals a sentence fragment, any sentence with an initial conjunction must be equally incorrect.

  9. FW said,

    November 7, 2009 @ 12:28 pm

    'Although' is a subordinating conjunction – I think Tom L Waters's beef is with coordinators, TB.

    Thanks, Mark!

  10. Layra said,

    November 7, 2009 @ 12:38 pm

    With regards to the use of sentence-initial coordinators and education, when I was in university, I and several friends noted a prevalence of starting sentences, even conversations, with "So". Instead of going up to someone and saying, "I had an idea…", the phrasing would be "So I had an idea…", "So I was talking to this guy the other day…" and so on.
    Our theory was that by starting a conversation with "So", the speaker was attempting to establish relevance, implying that whatever was said before is the first half of the coordination, since a conversation started with "So" only has the latter half.
    Although guilty of starting many a dialog with "So", I'd still find the idea of conversations started with "And" or "But" to be terribly bizarre.

  11. uberVU - social comments said,

    November 7, 2009 @ 12:39 pm

    Social comments and analytics for this post…

    This post was mentioned on Twitter by interests: Language Log: Also, check the back seat

  12. thomas said,

    November 7, 2009 @ 1:09 pm


    Seamus Heaney's translation of Beowulf starts with "So", an idiom that he says he got from his father's family. His explanation is more or less the opposite of yours — 'in that idiom "so" operates as an expression which obliterates all previous discourse and narrative, and at the same time functions as an exclamation calling for immediate attention'.

    I think this reinforces the point that this isn't a grammar issue, and it's not even a single issue of usage, but a case-by-case issue of style.

  13. Richard Thomas said,

    November 7, 2009 @ 2:04 pm

    This subject has even been debated in the UK Parliament. The drafter of the Finance Bill that contains the tax legislation giving effect to the Budget each year is quite happy to use both "And" and "But" to start a sentence. But there is a prescriptivist lawyer MP (Rob Marris) on the Committee examining the Bill who is completely against the usage. The report of the Committee looking at the 2007 Finance Bill contains this exchange

    Rob Marris: In relation to the amendments and to the debate on the schedule, may I again gently point out to my hon. Friend the horrible drafting? In amendment No. 68, proposed new paragraph 10A(5) starts with “But”. Similarly, in amendment No. 74, proposed new paragraph 15(7) of schedule 5 starts with “But”. In the schedule itself, paragraphs 2(3), 9(3), 11(5), 12(5), 14(3), 15(5) and 17(3) all start with “But”. Please could he have a word with the draftspersons and have them stop using that word? In almost all those cases, as far as I can tell, it is not only grammatically erroneous, but completely redundant.

    Kitty Ussher (Burnley) (Lab): Perhaps this is not a crucial point, but is my hon. Friend aware that The Economist style guide authorises the use of “But” at the beginning of a sentence?

    Rob Marris: I shall not engage in a long discussion. I am aware that many journalists use it; it does not make it right.

  14. Simon Cauchi said,

    November 7, 2009 @ 2:06 pm

    It's not just Seamus Heaney. There are lots of sentences beginning with "So" in Evelyn Waugh's 1932 short story "Cruise", subtitled "Letters from a young lady of leisure".

  15. Tom said,

    November 7, 2009 @ 3:22 pm

    For the record, Heaney follows his "So" with a full stop:

    So. The Spear-Danes in days gone by
    and the kings who ruled them had courage and greatness.

    I'd be curious to hear from any Anglo-Saxonists how literally that renders the original. (In Heaney's facing-page edition, there's no punctuation after the "Hwaet," but I've no idea what old-English punctuation conventions are.)

    I'm as inclined as the next guy to call NIC a dumb rule, but I do have some sympathy for my second grade teacher who taught it to me. I and my classmates were liable to start just about every sentence with an and or a but, and I doubt anyone here would like to read much prose like that. Perhaps NIC is a rule worth learning and then unlearning?

  16. Philip TAYLOR said,

    November 7, 2009 @ 3:22 pm

    (So) what is the function of Blake's leading "And" at the commencement of "Jerusalem" :

    And did those feet in ancient time
    Walk upon England’s mountains green:

    I am more than happy to accept that a conjunction may be used to link two sentences, and thus that any sentence apart from the first may therefore commence with such a conjunction, but to what does a conjunction such as "And" link when it appears at the commencement of the very first sentence ?

    [(myl) Well, in the first place, it wasn't originally the very first sentence. That verse appears in the preface to Blake's large poem Milton, following these two prose paragraphs:

    The Stolen and Perverted Writings of Homer & Ovid: of Plato & Cicero. which all Men ought to contemn: are set up by artifice against the Sublime of the Bible. but when the New Age is at leisure to Pronounce: all will be set right: & those Grand Works of the more ancient & consciously & professedly Inspired Men, will hold their proper rank, & the Daughters of Memory shall become the Daughters of Inspiration. Shakspeare & Milton were both curbd by the general malady & infection from the silly Greek & Latin slaves of the Sword.

    Rouze up O Young Men of the New Age! set your foreheads against the ignorant Hirelings! For we have Hirelings in the Camp, the Court & the University: who would if they could, for ever depress Mental & prolong Corporeal War. Painters! on you I call! Sculptors! Architects! Suffer not the fash[i]onable Fools to depress your powers by the prices they pretend to give for contemptible works or the expensive advertizing boasts that they make of such works; believe Christ & his Apostles that there is a Class of Men whose whole delight is in Destroying. We do not want either Greek or Roman Models if we are but just & true to our own Imaginations, those Worlds of Eternity in which we shall live for ever; in Jesus our Lord.

    Since he ends the prose passage with a reference to Jesus, the continuation "And did those feet [i.e. the feet of Jesus] in ancient times / Walk upon Englands mountains green".

    In the second place, he may (despite his opposition to the Stolen and Perverted works of Ovid and Cicero) be using an English copy of the Latin pattern, which was to conjoin phrases by placing et "and" in front of each of them, including the first: thus "et A et B et C", not "A et B et C". This was a fairly common feature of highfalutin poetic style in the 18th and 19th centuries.]

  17. John Cowan said,

    November 7, 2009 @ 5:22 pm

    Rob Marris gets the thing right, but he under-generalizes. He should be saying "I am aware that many journalists, professors, authors of great literature, and plumbers use it; it [sic; that?] does not make it right." In other words, some people believe that the "grammar rules" they exhort others to obey are truly from Mt. Sinai (though they would repudiate the connection).

  18. Simon Cauchi said,

    November 7, 2009 @ 7:34 pm

    Having just listened to Agathe's lovely cavatina "Und ob die Wolke" from Der Freischütz, sung beautifully on YouTube by the late Elizabeth Parcells, I'm sceptical of any Latin influence on Blake's initial "And" in "Jerusalem".

  19. Chris Waigl said,

    November 7, 2009 @ 8:11 pm

    Virginia Woolf's "A room of one's own" starts with "but", and has three sentence-initial "but"s in the first paragraph, used effectively in my opinion.

    (This passage, and specifically the initial "but", are used as an example in one of the textbooks, used by me in 2002, for analyse grammaticale et linguistique portion (the exact title varies) of the French CAPES/Agrégation exams for English. Unfortunately I do not have this book to hand at the moment.)

  20. Matthew Kehrt said,

    November 7, 2009 @ 9:24 pm

    I find the use of sentence-initial "and" to be stylistically disconcerting. However, since joining academia, I have become aware of the glorious word "moreover", which completely solves this problem (and which tends to be extremely overused in computer science academic writing).

  21. Dan Lufkin said,

    November 7, 2009 @ 9:46 pm

    I wonder whether Blake used "and" in And did those feet in ancient time/Walk upon England’s mountains green: in the way "et" is used in Et in Arcadia ego or Timeo Danaos et dona ferentes — to indicate the idea of "even." Jesus may have even walked in England.

    I was once gonna write "Timmy O'Donohue + Donna Ferentes" in a heart on a wall in Harvard Yard, but decided it would be a waste of chalk.

  22. John Cowan said,

    November 8, 2009 @ 5:37 am

    In addition, Blake needed an unstressed syllable at the start of the Jerusalem hymn, and "And" would do as well as any.

  23. Philip TAYLOR said,

    November 8, 2009 @ 9:18 am

    I have just remembered that my school song also had a verse that commenced with a conjunction (or perhaps I should say, disjunction) :

    Or lost, or won,
    The games we play
    Will stir our hearts
    For many a day.
    Then gather ye sons of Colfe around (etc.).

  24. Liz said,

    November 8, 2009 @ 10:01 am

    "nice to see that the "students need our insane rules" argument so early in the thread."

    Is this a common and easily sneered at position then? I'd love to be directed to other instances.

    [(myl) Two common variants of this argument have been given convenient acronyms by Arnold Zwicky: Zero Tolerance 1 (ZT-1) "If they do it too much, they should be told not to do it at all", and Zero Tolerance 2 (ZT-2) "If doing it sometimes gets them in trouble, they should be told not to do it at all". There are other subcases of the "students need our incorrect rules" meme as well, but perhaps these will do as a start.]

    And I wasn't saying that students NEED them. Nor do I see it as a "rule". Just that if, after ten years of compulsory education a student is incapable of writing an acceptable sentence (acceptable in a mainstream world, not the highly sophisticated world of compartitive linguistics) you use anything that works.

    [(myl) It's past time, it seems to me, for a program of research to support evidence-based writing instruction. I've previously argued for some research on the uptake side (e.g. "Prescriptivist science", or "A test kitchen for stylistic recipes"). But we also need research on what instructional methods actually work. There's not much useful and credible research of this kind out there, at least not much that I've been able to find.]

    As for Anglo Saxon punctuation, the blurring of the line between oral and written cultures might mean they were a bit – flexible? Fluid? Both Hwaet and So work as "Hey you. Pay attention".

  25. Dan Lufkin said,

    November 8, 2009 @ 12:13 pm

    @ Philip Taylor — In Latin, Italian, Dutch (& Afrikaans), French and Spanish, at least, the or … or … construction is the standard way of saying "whether … or …" "Whether (we) lost or won," in your song. (It don't make no never-mind.)

    And (harking back to myl's note) a leading Et … et … was (and still is) the standard Latin way of saying "Both … and… ."

  26. Nick Lamb said,

    November 8, 2009 @ 3:13 pm

    I'd assumed that Blake intends this poem to work fine without the prior paragraph. If it's OK to begin a story in the middle (audiences seem to take this in their stride) then it's no problem to leave mention of which feet we're talking about for a few lines. As lyrics for a pop song I think a similar construction would go unremarked

    And now, the end is here
    And so I face the final curtain

    Oh, it turns out it already was and it already did.

  27. Dan T. said,

    November 8, 2009 @ 3:14 pm

    What's the context in the British finance bill that causes so many paragraphs to start with "But"? What are they conjoined to in this case?

  28. Paul said,

    November 9, 2009 @ 9:04 am

    Just had a piece of writing peer-reviewed by a linguist who complained about an "incomplete sentence" in my text. I looked to see what it was. And it was an example of NIC. :-(

  29. Ken Brown said,

    November 9, 2009 @ 9:21 am

    "I wonder whether Blake used "and" in And did those feet in ancient time/Walk upon England’s mountains green: in the way "et" is used in Et in Arcadia ego or Timeo Danaos et dona ferentes — to indicate the idea of "even." Jesus may have even walked in England."

    I don't think so. Blake is referring to the Joseph of Arimathea/Glastonbury legend but he is not suggesting it is historical fact. The opposite – its a rhetorical question that expects the answer "no". The whole verse is a series of such questions. The point is that Jerusalem was NOT builded here, so we have to do it ourselves. "…till *we* have built Jerusalem, in England's green and pleasant land"

  30. Daniel Smith said,

    November 9, 2009 @ 3:30 pm

    I don't see a good email address to make a suggestion, so I'm hijacking this comment thread to give an idea for a future post: I've been told I use too many "to be" verbs in my writing. Do good writers really avoid "to be" verbs, or is this like some of the other proscriptions you guys like to bash on here?


  31. Jair said,

    November 10, 2009 @ 1:54 am

    I’m pretty sure that if LL blogs about that topic they will be required to make the subject line a certain very obvious pun.

  32. Richard Thomas said,

    November 10, 2009 @ 12:18 pm

    In response to Dan T, the normal rule is that each legislative unit contains one sentence. That sentence may be a general rule, with any exceptions set out afterwards. For example, one of the provisions that Rob Marris refers to reads

    (2) The amendment made by sub-paragraph (1) has effect in relation to accounting periods ending on or after 6th March 2007.
    (3) But income which arises in an accounting period beginning before that date is to be chargeable to corporation tax as a result of that amendment only if it arises on or after that date.

    In the past the exception would often be in a separate sentence in the same provision and would start "Provided that…..". Later it might be that the general rule would start "Subject to sub-paragraph (2), the amendment made by….", with sub-paragraph (2) describing the exception.

    But now the drafter just says "But". They also say "And" (and "Instead") at the beginning of sentences.

    I also happen to have an Act of 1797 before me. In that every single sentence starts with either "And" or "Provided…that

RSS feed for comments on this post