Monday 18 July 2011

On identifying phishermen

Correspondents (of the radio kind) have been keeping the phone hot this week in the wake of a report claiming that spelling mistakes on websites can cut online sales by half. I'm not surprised. If website writers don't take the trouble to satisfy the norms of standard English - which is defined chiefly by its spelling, punctuation, and grammar - then they must expect to encounter trouble. People are very ready to make deductions about the background of a user based on language use, and the argument 'carelessness in spelling must mean carelessness generally (and thus an unsatisfactory product)' is applied regardless of the realities. Quite clearly, firms need to employ proof-readers if they sense they have a deficiency in the spelling department. There are plenty of free-lancers out there willing to help.

Interesting research questions still need to be answered. What are the areas of internet activity that generate these expectations? Clearly there are some outputs where deviations from standard English are normal, expected, and valued. And what pragmatic effects does nonstandard usage on the internet convey? One point which didn't get a mention in the BBC report is the way nonstandard English can be an important clue to the dubious origins of a message. Here are three examples of phishing that I received recently, all from someone purporting to be a gmail service provider and wanting my personal details. The nonstandard English provides the clues (some of which I italicize below). There are pointers of spelling, punctuation, and grammar, as well as awkwardness of style and inconsistency (eg in the use of capital letters). Probably the whole of the first example should be in italics, given the blend of sentence structures!

(1) 'We make every effort to ensure that we provide the Ultimate Security required for maximum protection because we are detecting unusual activity on some user account, we have decided to protect each account with a user account control to protect user privacy and make sure each user account is not accessed unauthorised.'

(2) 'We have received several complaints from users unable to gain access to their email account, as a result of that, we are upgrading our security systems and making sure each user account is not accessed unauthorised. We do not want you to loose access to your Account since your login information are no longer valid on our database system. Now, the Gmail Account Team need to confirm your profile details below for verification purpose and to confirm that you own this Account.

NOTE: If you would like to continue using our services, please click on the reply button and email us the afore mention details immediately for confirmation and validation. We apologize for any inconveniences. Thanks for Using our Service.'

(3) 'This is an important information regarding your Google account. We have just realized that your account information on our database system is out of date, as a result of that we request that you to verify your Information by filling your account information below.'

As time passes, and people become increasingly experienced in reading and interpreting web pages, they are developing intuitions about the status of the originators. This applies as much to matters of graphic design and choice of style as to content. What we are seeing in these examples is the emerging role of nonstandard English as an index of internet illegitimacy. I expect the same sort of thing takes place in other languages? Examples welcome.

Friday 15 July 2011

On enquiring about inquiry

A correspondent reports something he was reading in The Times this week:

1356 BST: Jemima Khan has a complaint about the police investigation into phone hacking. 'Not much hope for hacking Inquiry when they can't even spell it... I received "Operation Weeting Enquiry [sic] Questionnaire" last week,' she tweets.

He notices that I often write in my blog about enquiries from correspondents. He concludes: 'What are the rules and/or usage in British English? I know that in American English it is always inquiry.'

Well, not always, actually. The Cambridge Corpus of American English shows a preference of 97% for inquire and 88% for inquiry. That's a dominant usage, certainly, but not a universal one. In Britain, the picture is extremely mixed. The British National Corpus shows almost exactly twice as many enquire as inquire, and twice as many inquiry as enquiry. The world picture, amalgamating different spelling traditions, is mixed too. Google shows inquire five times more common than enquire, and enquiry seven times more common than inquiry. But all four forms are frequent. Not surprisingly, then, most dictionaries throw in the towel and say the i- and e- forms are interchangeable. The OED, for example, simply lists them as alternatives, but adds a note under enquire:

'An alternative form of inquire v. The mod. Dicts. give inquire as the standard form, but enquire is still very frequently used, esp. in the sense "to ask a question".'

Could there be a sense difference? Prescriptive grammarians tried to find one, citing the difference between insure and ensure as justification, and their view did have some influence. The i- forms should refer to impersonal, formal investigations, it was recommended, whereas the e- forms should be used only for personal questions, and doubtless many people tried to make their usage conform to this distinction. The lists of examples in the large corpora, however, show many counter-examples, and it is difficult to resist the conclusion that, for most people, the forms are as interchangeable as judgment and judgement - in other words, influenced by such factors as region, house style, and institutional preference, but not by anything semantic.

The original forms in English were with e- (from French enquerre), but in the late 14th century we see i- spellings appearing, as people tried to reflect the Latin origin of the words (inquirere) - a common practice at the time. Dr Johnson put all his weight behind the i- forms in his dictionary, and doesn't include the e- forms at all. Modern style guides seem to be going the same way. The Cambridge Guide to English Usage (2004), for example, concludes thus:

'Given no consistent ways of differentiating the two spellings, and the fact that differentiation is unnecessary, it makes sense to consolidate the use of one or the other. Inquire and inquiry recommend themselves as the spellings made first among equals by the Oxford Dictionary, and the fact that they are strongly preferred in North America.'

That doesn't make the e- forms wrong, of course, as the quotation from my correspondent suggested. And it's nonsense to suggest there might be a correlation between this choice of spelling and the conduct of a policy enquiry.

Tuesday 5 July 2011

On mouth-filled speech

A correspondent writes with an enquiry that needs to be quoted in full:

'This morning I tried simultaneously to brush my teeth and talk.  I tried saying, 'I don’t know,' and the listener managed to understand my muffled 'words'.  Actually, they could be thought less of words and more as pulsated approximations of words, three throbs with the first one neutral, the second a bit higher, and the third ending on a lilt.  Since the words 'I don’t know' are used so often in English, it wasn’t difficult for my listener to guess what I meant.  And that got me thinking, how much does this sort of 'speech'—hummed, or pulsated approximations of real words— factor into the English language, as well as others?  I imagine that for any language, the most common words and phrases would, even if intonated in such a 'muddy' manner, still be understood because of their familiarity and frequency of use.  Is this sort of speech ever used for histrionic or comic effect?  Or have any authors ever exploited it for inventive literary purposes?'
This is an area which, in phonetics, would fall under the heading of paralinguistics - though I have to say mouth-filled speech isn't one of the categories recognized when Quirk and I first studied vocal effects back in the 60s. It just didn't turn up in the corpus - unsurprising, really, as 'Don't talk with your mouth full' is a (?universal) pragmatic prohibition that we learn from our parents at around age 3, and the recordings of relatively formal situations we were using then simply didn't present the relevant situations. The surreptitious recording of bathroom or dining-room speech wasn't a top priority at the time.

It's more than just politeness that's at stake. There is a risk of choking. And unintelligibility. But etiquette is a dominant factor. Some people, if asked a question at exactly the point where they have taken a mouthful on board, simply refuse to speak until they have swallowed, which can produce an awkward silence in the conversation (though the mouth-filled one will usually use facial expression or hand gesture to explain what's happening). Listeners understand the problem if they've been brought up in that way. (I muse over my parenthesis above. Is it etiquette in all languages? It is in all the language situations I've experienced.)

Despite the lack of examples in corpora, mouth-filled speech is really rather common. I suspect most people do it, from time to time, in informal eating situations, when they feel the urgent need to make a point. And eating is only one of the relevant situations. Other examples, in addition to speaking while brushing the teeth, are

- speaking while holding a writing implement in the mouth (while the hands are otherwise engaged), as I've often seen in business meetings
- speaking (or trying to) when the dentist, just having filled your mouth with implements, asks you if you had a nice holiday
- and relatedly, speaking after having had your gums filled with anaesthetic
- speaking with pins in the mouth, while sewing
- speaking with a pipe or cigarette in the mouth
- speaking with a hand or finger in the mouth, sucking it better after a hurt
- speaking with ill-fitting false teeth
- little (and sometimes not-so-little) children, sometimes try to speak while keeping a dummy (pacifier) in the mouth
- speaking with a decorative item in the mouth, such as a pierced tongue
- for boxers, speaking with a gum-shield
- in old-style elocution, speaking with a pebble in the mouth to improve one's pronunciation - a technique supposedly used by Demosthenes to overcome a stammer
- more dramatically, movies regularly show us someone trying to speak with a gag in the mouth
- or talking while someone else is in their mouth, as with a passionate kissing scene.

These situations are common enough to have made me role-play mouth-filled speech in listening comprehension exercises, when I used to do some EFL teaching in summer schools. Solo, I hasten to add, in view of the last example.

Linguists are well aware of the importance of avoiding situations where something interferes with natural speech production. Field linguists watch out for any physical limitations in their informants - it would be unwise, for example, to rely greatly on the phonology produced by an aged speaker who had lost all his teeth. And some of the semiotic transcriptions of body behaviour from the 1960s include symbols for such effects as 'speaking through clenched teeth', 'speaking while licking one's lips', and 'speaking with mouth pursed'. However, these are just general markers. I don't know of any phonetic descriptions at the level of the segment.

Do authors do it? I haven't come across any. They seem to leave the effect to the reader's imagination. Here's J. M. Barrie in A Widow in Thrums (Chapter 3):

' "Ye daur to speak aboot openin' the door, an' you sic a mess!" cried Jess, with pins in her mouth.'

The character has that accent throughout; no special effort is made to represent the effect of the pin-holding. Here's George Eliot, in Scenes of Clerical Life (Chapter 1):

' "So," said Mr. Pilgrim, with his mouth only half empty of muffin, "you had a row in Shepperton Church last Sunday." '

That sentence would certainly have sounded differently. And even Charles Dickens, so good at depicting the idiosyncrasies in an individual's speech, leaves this effect to the reader, as in Nicholas Nickleby (Chapter 5):

' "This is the way we inculcate strength of mind, Mr Nickleby," said the schoolmaster, turning to Nicholas, and speaking with his mouth very full of beef and toast.'

A rare example of an author trying to represent the segmental phonetics of mouth-filled speech is Anna Pickard in The Guardian (27 April 2006) which begins:

' "Fankky, i's ow-wajus. I fine i' affo-uuti owajus. Va figiss ... hangom, suwee, nee to swa-oh." Frankly, it's outrageous...'

And she goes on:

'And what, I ask, is so wrong with talking with your mouth full? In an age where multitasking is a marketable skill, surely the ability to eat and keep up your end of the conversation at the same time should be positively commended. '

She specifies three benefits:

'Time management
There simply isn't time in the day to set aside a separate amount for eating and for talking. By combining the two activities, an incredible amount of time can be saved. Also, none of your companions will ever need to ask what you had for lunch again. They will know, because they can see.

Portion control 
The process of eating while talking can do wonders for the figure. Anatomically speaking, the act of sucking in air for the talking while holding food in the oratory position should, in theory, bring more air into the food, thus inflating it, and making you feel more full (if slightly gassy). While this hasn't been scientifically proven as far as I know, speaking as a university graduate, it certainly sounds like a convincing theory. My degree is in dramaturgy.

By the simple act of talking while eating, you can easily ensure that you will be memorable to everyone you meet. While what you were saying might have been otherwise forgettable, no one will ever forget you if you gave them a good eyeful of bolognese while you were saying it.'

It's nice to have the opportunity of resurrecting this piece from the journalistic past.

If readers of this post have come across any other examples of mouth-filled speech, especially in literature and in languages other than English, I'd love to know of them, as I'm sure would my correspondent.

Monday 4 July 2011

On texted vs texed

A correspondent writes to ask about the past tense of the verb to text. He uses texted but is aware that many people say text, as in She text me yesterday. 'Why is this>' he asks. 'Is it something to do with the consonant cluster at the end being difficult to pronounce?'
The historical situation is clear in the OED. When text became a verb in English, back in the 16th century, meaning 'write, inscribe', it had the expected regular past tense form, -ed. We find an early use in Much Ado About Nothing, when Claudio says 'Yea, and text underneath, here dwells Benedick the married man'. And we find a past tense in Thomas Dekker's play Whore of Babylon, 'Vows have I writ so deep... So texted them in characters capital...' That's 1607.

Unsurprisingly, then, when text became a verb again in the 1990s, in the modern sense, it followed the normal pattern, and texted is the form given in all the dictionaries. So the interesting question is, why has an alternative form developed. It's very unusual to find a new irregular past tense form in standard English. It does happen, as we see with the preference for shorter broadcast and forecast alongside broadcasted and forecasted, but that was influenced by the basic verb cast, past tense cast. We don't have the same situation with text.

Pronunciation is probably part of the answer. There's nothing intrinsically difficult about the consonant cluster at the end of text, as we don't have a problem with other words in English which have exactly the same consonant cluster in that position in a word, such as next, vexed, faxed, boxed, sexed. Indeed, there is evidence from the history of English that the 'xt' pronunciation is actually easier than some alternatives, as when we see asked change to axed in many regional dialects. But adding an -ed ending alters the pronunciation dynamic. We now have two /t/ sounds in a rapid sequence, as we had in broadcasted, and that could motivate people to drop the ending. Speakers generally prefer shorter forms.

This then means that we have a present tense and past tense which aren't different, but that's nothing unusual in English, as we see with bet, bid, burst put, and others. Indeed, text as a past tense has something going for it: it actually sounds as if there is a past tense -ed form there already. Compare the sound of I fix, mix, fax, sex (meaning 'decide the sex of', as in The vet sexed the kittens), and so on, which in the past tense are fixed, mixed, faxed, sexed. Text sounds like them, and even though there is no verb tex, the pronunciation analogy could still operate. (Also, of course, in colloquial speech, text is often pronounced /teks/ anyway.) So maybe people are beginning to think of text as if it were texed.

Whatever the reasons, we do now find forms such as texed and tex'd being used with increasing frequency. I think it's only a matter of time before we find it being treated like broadcast in dictionaries, and given two forms.