Ian Preston, a London economist, did a bit of research of his own into the issue of the police officer who has been accused of having a little passive-aggressive fun by peppering his inquest evidence with song titles. "It seems to me," Ian remarks in a classically British understated way, "that the evidence cited on this in newspaper discussion is a little underwhelming."
It sure is. Ian not only found yet more song titles in the same police testimony; he then undertook the experiment of checking another random text for comparison, and found song titles there too. What's more, the second text he took was an email on an entirely non-song-related topic from a professional grammarian. What's more, the professional grammarian was me.
I quote in extenso from Ian's email (with his permission):
The songs which they cite are: Point of no return — credited to Duran Duran (BBC), Phantom of the Opera (ES, Telegraph), Buzzcocks (Guardian); Enough is enough — Barbara Streisand (BBC), Alice Cooper (ES), Donna Summer (Guardian), Jamie Lidell (Telegraph); Line of fire — The Journey (BBC, ES, Guardian); First time — Robin Beck (BBC); Self Preservation — from The Italian Job (ES, Telegraph) Blinding Light — August Burns Red (ES, Telegraph) Fuck My Old Boots — The Membranes (ES, Guardian, Telegraph) I'm Kicking Myself — As Tall As Lions (ES, Guardian, Telegraph) Quiet Moments — Chris de Burgh (ES, Guardian, Telegraph) Any Other Way — Backstreet Boys (ES, Telegraph) People Talking — Jay Z (ES).
More or less all of these are common and, in the context, appropriate short phrases. In one instance the four reports find four different singers for one title (and in another three), showing just how common the phrase is. (There are also other uncited versions of "Any Other Way" for instance - eg Celine Dion, William Bell - and "First Time" - eg U2, Styx, Lighthouse, . . .) In one case the phrase is not even a proper song title, just a phrase used in a song in a well-known film. The only phrase that strikes me as slightly unusual is "Fuck My Old Boots" but here the song cited is hardly well known and the phrase plainly predates the song (and are we meant to believe the same officer is an aficionado of both The Membranes and Chris de Burgh?). (To be honest, the fact that a police officer finds this an appropriate way to express himself at an inquest into someone he is alleged to have killed shocks me more than the thought he might have played a game with song titles in his evidence.)
Looking at the quotes from the testimony in the Guardian I found at least five other song titles by idle Googling: "Play It Back" (Lonnie Smith), "In My Mind" (Heather Headley, Antiloop), "Towards Me" (Sam Brown), "On the Trigger" (Bidwell), "Turn My Back" (Sick of it All). I wouldn't be surprised if there are more but what does it prove?
Just for a comparative example, as I thought of emailing you about it, I also checked the text of your brief email to me from this June.
I remember that email. It was about an observation of his concerning the problem of getting used to in plain form constructions — a topic that Mark Liberman coincidentally addressed just the other day in "Didn't use(d) to be". What I wrote to Ian was this:
Spot on. You have noticed a strange little feature of English that I think is quite revealing. "Used" is MORPHOLOGICALLY a preterite; but for "Didn't he ___ to be rubbish?" we need a plain form to fill the blank. "Used" normally only appears in its apparently preterite form. If you try to force it to act as a plain form, you get unquestionably ungrammatical strings:
*In order to be eligible you must used to be in the RAF.
*I enjoy going to the weddings of my old girlfriends; it gives me a thrill to used to have sex with the bride.
Surely nobody would write those. And yet if we can't have "use(d)" after auxiliary "do", we have no interrogative corresponding to "He used to be rubbish"! So people often bite the bullet and say either "didn't he use to" or "didn't he used to".
What hardly anyone can bear to conclude from all this is what I think is actually correct: there is NO ANSWER to what is correct here. The language is not well designed enough to provide an answer. There is an embarrassing inconsistency in the syntactic regulations, very much like an embarrassing inconsistency in a legal or regulatory system. And we just sort of vamp. Nobody knows what to do; everybody improvises with what we've got.
Of course, I'm anthropomorphising: this is nearly always unconscious. But in rare cases there are people like you who notice it explicitly.
Well, here is Ian's account of his experiment on my innocent prose (where I would have said you didn't have a snowflake's chance in hell of finding any buried song titles):
I fairly quickly found what appear to be five song titles: "Fill the Blank" (Dang Show), "Old Girlfriends" (John Wesley Harding), "Bite the Bullet" (Motorhead, Roadstar, Gillan, Machine Head, …), "No Answer" (Falco, Juliana Hatfield, Watermarks), "Nobody Knows" (Pink, Nik Kershaw, Tony Rich) Most of these are quite obscure but then some of the songs cited for AZ8 are also obscure. I don't suppose that you were making any effort to insert song titles into your reply but then perhaps the incidence and obscurity of these examples doesn't match those in his testimony.
I guess that this is difficult to check systematically. Can ordinary language corpora be cross-referenced against song catalogues? Maybe he has admitted it or boasted about it to colleagues but, if not, is there any way of assembling evidence that speaks to whether or not some of the reporting is or is not excessively credulous?
I think it is clear what the answer to this must be, don't you? It seems almost certain that the newspapers have been excessively credulous. There is but a snowflake's chance in hell of establishing that the presence of the allegedly song-derived phrases could not be due to chance. Unless the officer known as AZ8 boasted in front of more than one witness that he had mocked the inquest process by sticking song titles into his testimony, they cannot possibly substantiate the charge against him. If Language Log writers (or for that matter linguistically interested diligent economists) are called as expert witnesses, they are going to have to testify for the case of the defense.