McWhorter on Initial So.

I have often expressed a combination of irritation and admiration when it comes to John H. McWhorter, and so it is now. For quite some time I have been wondering about the current popularity of starting sentences with “So” and wishing someone would explain it, and he has done so about as well as could be hoped for. But he has done it in a podcast, a format I dislike for its inefficiency, and it takes him over half an hour to make a point that should take a couple of minutes. After nine minutes of blathering about other uses that are clearly not what people have been pestering him about, he gets to the one that matters, the “Terry Gross” one that is not motivated by a change of subject or the like. After sixteen minutes he finally gets to the explanation: it’s a replacement for sentence-initial “well” (which is what has been familiar to me all my life), and that is itself a replacement (after many centuries) for Old English hwæt ‘what,’ used in a similar way. Why the new forms? Because language changes, and you can’t predict how. That’s good enough for me! (We discussed the issue a few years ago here.)


  1. David Eddyshaw says

    Amen re podcasts, standing proof that just because something is possible, it doesn’t mean it’s in any way a good idea.

    I’ll warm to podcasts when they invent a practical way of skim-reading them. Those Sumerian accountants, they built better than they knew.

    [As the resident McWhorter-supporter, I would like to comment positively on his podcast – but life is too short. Sixteen minutes!]

  2. Two words – accelerated playback.

  3. I think the point of podcasts is you can listen to them while doing dishes or whatever. I’ve never managed to get into the habit though.

  4. Your car’s speaker system is your friend.

    Put the time wasted in a traffic listening to stupid music or dumb talk radio to some productive use.

  5. David Marjanović says

    If it’s American enough, it might be a translation of German also.

  6. Just as a heads-up, I’m having computer problems and may not be able to post for a couple of days (if I have to order a new computer via Amazon, that’s how long it will take to be delivered), so if that happens, you’ll know what happened and can talk among yourselves.

  7. blathering about other uses that are clearly not what…

    I hate so halfway through a sentence as an implied but undemonstrated “therefore” or consequence of the first half, as used by politicians. ‘These are dangerous times, so we believe it’s the prime minister who should apologise.’

  8. talk among yourselves

    Are you sure you don’t just need an extra hard disc? They’re much cheaper than a whole computer.

  9. languagehat says

    Well (or “so”),I got this computer in 2011, so it’s given about all it has to give.

  10. So it probably just needs a dust & wash. Shake the crumbs out of the keyboard, feed a family of five.

  11. His philistinism regarding /Beowulf/ results in missing out on the perfect closer: the most celebrated (and contested) translation of the last 20 years begins, “So. The Spear-Danes in days gone by…”

  12. J.W. Brewer says

    I would classify myself as a generally McWhorter-supportive resident, although I am happy to defer to David Eddyshaw’s chieftanship of that faction. And definitely not so supportive as to listen to a lengthy podcast in search of some specific tidbit that I could just scroll down to quickly if a transcript were posted somewhere. I of course have the life parallels to wrestle with. (McWhorter and I are exactly the same age and grew up maybe 25 miles away from each other. When I was thirteen I somehow managed to fulfill the letter but not the spirit of a school assignment by writing about the variousness of languages spoken in Surinam — yet he was the one who grew up to “turn pro” and actually make a living studying language use in Suriname, as it had been by then respelled.)

  13. It’s not really the Terry Gross ‘so’ that people are interested in or bothered by. It’s the Mark Zuckerberg ‘so’ (this is all at roughly minute 11), the kind that introduces an explanation. And in that usage, as David says, it’s similar to (a copy of) German also.

    That’s all this McWhorter needed to say. And another thing: I can’t stand Blossom Dearie.

  14. Suriname, as it had been by then respelled.

    I remember when I was working as an editor on the Price Waterhouse series of Doing Business in Other Countries I was very grumpy at this change, and I sent around a memo saying that if we were going to use the new spelling we should also start saying “Suri-NAAM-e” instead of the normal English pronunciation.

  15. squiff-marie von bladet says

    I sent around a memo saying that if we were going to use the new spelling we should also start saying “Suri-NAAM-e” instead of the normal English pronunciation.

    So you should, he soed. I certainly do.

  16. I first started noticing “so” as a sentence opener in NPR interviews between five and ten years ago although I’m sure it predates that time. What struck me was that it was most used by academics of various persuasions being interviewed about topics in their particular expertise. Always when answering a question. So, I posit that this has its origins in academic discourse at colleges and universities, becoming widespread sometime in the oughts.

  17. I’ve always taken initial ‘so’ to be following an omitted summary of the current situation, which is unnecessary because it is obvious to everyone present.

    The audience is in their seats, and the featured speaker is on the stage, so it’s time to begin the presentation by introducing …”

    We’re sitting in this radio studio, and the On Air light has just come on, so let me begin by welcoming our guest …”

    I don’t really remember when I started hearing it as a sort of marker that some kind of presentation, lecture or similar event was about to start, but I think it must have been in the 1980s if not earlier.

    When answering a question, it comes after an implied Given everything that’s been said thus far,

    Another, less formal, equivalent is “All right”.

    “All right, today I want to look at how we can use Ohm’s Law …”

  18. President Reagan wielded the word “Well” with power, bringing all he spoke to into his comfort circle. You might not know what he’d say, but you expected it to be on spot somehow. “Hwaet” and “So” might serve the same purpose.

  19. When answering a question, it comes after an implied Given everything that’s been said thus far,

    No it doesn’t, except in the vacuous sense that any sentence opener implies that. If the interviewer asks “How did your band get started?” and the response is “So we were all in this bar one evening…” how exactly does your implication work?

  20. Starting sentences with “So,…” has long been a stereotypical feature of Minnesota speech. The correspondingly stereotypical way of ending sentences would be with “… then.”

  21. I listen to podcasts during household chores, while redoing my nails, shopping for groceries and such. The History of English Podcast is quite well-done for a non-specialist, and The Saga Thing has a fun format (they summarize Icelandic sagas and then judge them on categories like “best nickname” or “best bloodshed”). I also listen to language courses, random things in languages I need practice, and themes outside of the scope of’s main interests.

  22. I think anyone who sees podcasts as just an inefficient way to extract information is rather missing the point, which is to spend time in the company of, hopefully, congenial people discussing something of interest to you. And yeah, I don’t think people typically just sit down and listen while starting at the wall. You’re usually doing chores or going for a walk or driving somewhere or play a videogame that doesn’t require your full attention or what not. Obviously, if you’re just looking for facts, a podcast is not the medium to use.

  23. J.W. Brewer says

    The defenses of podcasts point out their merits for consumption in a particular fashion (for people who like consuming things in that particular fashion), but those merits do not alleviate their shortcomings for secondary uses in contexts like the present one, where hat wants to point out and encourage discussion of a single point discussed somewhere in the middle. Perhaps sometime soon automated transcript-generating software will be so good that hat could turn the relevant bit he wanted to post about into cut-and-pastable text with barely more effort than it would take to cut-and-paste from an article, but I don’t think we’re there quite yet.

  24. The History of English Podcast

    Podcast on The History of English Podcast?

    Yeah, but, of course.

    Any other way would be disrespectful to this fine method of consuming information.

  25. Thank you! The more I read stories online the more I encounter this, and it’s been eating at me.

  26. The Log has a new post about this.

  27. I just heard a classic example of the new system; NPR’s Scott Simon interviewed Wolf Cukier (SOO-kee-er), the high school senior who discovered a new planet, and Cukier began his first response: “So, the cool thing about this planet is it’s a circumbinary planet. So what that means is that the planet orbits two stars at the same time. So if you think to Luke’s home planet in Star Wars…” Later: “How did you discover a planet?” “So, the test telescope observes the sky…” There are other examples; you can hear the whole (3-minute) interview here. I trust no one will claim that people began their answers to simple factual questions with “So,” let alone began a string of consecutive sentences that way, before relatively recently; I’m still hoping to learn more about when and where this started.

  28. @languagehat: As I said above, starting just about any sentence with, “So,” is standard stereotypical behavior for Minnesotans. I don’t have a ton of experience with people from Minnesota, but in my limited experience they can indeed insert, “So,” at the beginning of just about any utterance, including the answers to ordinary questions.

  29. But we’re not talking about Minnesotans, we’re talking about virtually the entire population of people who talk on the radio. I’ve even heard this automatic “So…” from Englishpersons. It’s clearly some sort of sweeping change, and I’ve just discovered an earlier discussion of it here from 2013, starting with PK’s comment. And Heaney’s irrelevant Beowulf usage is brought up there as well — I guess it’s so memorable it has to be mentioned.

  30. AJP Crown says

    It’s called ‘explanatory so’. It announces that an explanation is coming. There was an explaining woman on BBC tv News last night from a hospital in Indiana or Indianapolis or somewhere else Indian, a mudfud who has studied the coronavirus for 20 years, and she responded to literally every question with an initial “So…”

  31. It’s called ‘explanatory so’. It announces that an explanation is coming.

    I don’t believe it. It doesn’t only come before explanations, it comes before every response. If anything, it announces “I’m about to start talking.”

  32. Stu Clayton says

    I would call it a segue scam. It is used to create an impression that the sentence following it is being pulled by the current train of thought, an impression that the sentence connects reasonably with what has gone before.

    Every answer that begins with “so” is a pertinent response to the question. Every question that begins with “so” is pertinent to the previous answers and questions. Every remark that begins with “so” is a pertinent extension of previous remarks.

    So that’s what they want you to think anyhow.

  33. You’re way overinterpreting. This isn’t some Machiavellian plan, it’s an utterly unconscious speech habit. I repeat, this isn’t something the occasional Minnesotan or Machiavellian does, it’s very common and spreading like wildfire. It’s like the spread of discourse “like” a few decades back. I wish I could find a transcript of someone talking on one of these radio programs, so people who aren’t familiar with the phenomenon could see what it’s like.

  34. Stu Clayton says

    Overinterpreting ? I am merely describing a phenomenon. Of course it’s unconscious. The agent is the phenomenon, the speakers just go along for the ride.

    It has exactly the same effect as if it were deliberate – it creates an impression of continuity and attentiveness with minimal effort. It signals sociability.

    That’s why it appears to be “spreading like wildfire”. It’s unlikely to be the sum of individual, considered choice. You yourself call it unconscious. So it’s more like a coronavirus. It catches you, you don’t catch it. But you do anyway.

    By the way, I see discourse “like” in a similar way. It’s a phenomenon of automated self-deprecation. Like, it’s nice to be modest and deal with modest people, unlike those who don’t use “like” in this way.

  35. John Cowan says

    The defenses of podcasts point out their merits for consumption in a particular fashion

    Similarly, the defenses of wine-drinking point out its merits for discovering the chemical composition of the wine in a particular fashion. (This example due to Dennett in his article on qualia.)

  36. Stu Clayton says

    I take that to be Dennett’s sarcastic way of saying “if you’re just looking for facts, wine-drinking is not the medium to use.” [modified from upthread]

  37. Bathrobe says

    It’s called ‘explanatory so’. It announces that an explanation is coming.

    I don’t believe it.

    Well, then, it’s an extension of the old usage that “an explanation is coming”.

    Another one I’ve noticed in Australia is the use of “and yeah”. It seems to be used when people have said a few things, decide to add another one (“and”), and then realise they have nothing to say and make it into “and yeah”. Not as ubiquitous as “So” but it still stands out for me.

  38. Well, then, it’s an extension of the old usage that “an explanation is coming”.

    Well, everything is an extension of something else — bits of language are very rarely created ex nihilo (“blurb” being an exception). The tricky part, the part you usually need a linguistics course to take on board, is not letting the original sense affect your understanding of how it’s used now.

  39. John Cowan says
  40. AJP Crown says

    I may have caused a misunderstanding.
    There’s an old explanatory so:

    “Why are you carrying an umbrella?”
    “So I won’t get wet if it rains.”

    And a new explanatory so:

    “Why are you carrying an umbrella?”
    “So, I kept getting wet and then finally my mother bought me one.”

    So what I’m talking about is the new explanatory so.

  41. David Marjanović says

    The new no-longer-explanatory one is exactly parallel to the use of also in German.

  42. Lars Mathiesen says

    And the Danish, exactly parallel, use of altså was peeved against 50 years ago. Nice to see that English is getting with the program.

  43. AJP Crown says

    German also & Skandi altså yes, right.

    It won’t satisfy Language.

  44. People who have experienced not-being-attended-to in conversations use ‘so’ to announce that they have something to say, attention please! I’m referring to women when outnumbered by men, shy people amongst the loud, and so forth. It’s a way to bring balance to the debate. I believe Elizabeth Warren picked up the habit as a Harvard professor who doesn’t look like a Harvard professor.

  45. So what I’m talking about is the new explanatory so.

    But what I’m talking about is not explanatory. Here’s an imaginary conversation:

    First, tell us who you are.

    So, my name is Jon Jonsson.

    And how did you get here today, Jon?

    So, I took the 1 to Times Square and walked from there.

    Tell us about your discovery.

    So, I was in the lab…

    This is what those interviews sound like. There is nothing explanatory, and it is not about trying to bring balance to the debate. It’s just a verbal tic. I know y’all are trying to make sense of it in terms that make sense to you, but you don’t actually know what I’m talking about. You need to listen to NPR.

  46. @languagehat: I do notice that you made your hypothetical “So” user of Nordic descent—although he apparently comes from Wisconsin, not Minnesota.

  47. Yes, he works in the lumberyard there.

  48. John Cowan says
  49. I think this might amuse David Eddyshaw and maybe others too. So I am listening to a lecture on the European part of the IE family from McWhorter’s course “Language Families of the World” on Great Courses (never mind, I am tired of the primary and coronavirus takes and need some distraction while grass grows, I mean, code runs) and he says this (para)phrase:
    “Celts originated somewhere in the Caucuses. Part of them moved east and became mummies and another part moved west and became Welsh.”

  50. David Eddyshaw says

    Indeed. Our genes span the entire Tarim-Liffey family.

    I expect the Hittites were secretly Irish, too. The evidence is there if you know where to look. (Mursilis/Murphy …) They were deeply influenced by the O’Hurrians culturally.

  51. What a mursilis pun!

Speak Your Mind