There's much to be said in favour of predictive text, the bit of software in your mobile that guesses the word you're aiming for so you don't have to spell it out letter by letter. It's much simpler now to form coherent sentences and phrases and not have to use awkward SMSese expressions such as "l8r". A shopping list, a rendezvous switch, a billet doux, all have been rendered easier by predictive text.
A limit, however, is placed by the size of the dictionary in the phone. And this is where you notice oddities creeping in. Not just what the phone dictionary doesn't have, but what it does. My 2005 Nokia won't recognise the words "Mozart" or "Beethoven", for instance. But it does recognise "Wagner" and "Strauss". It is familiar with "Nazi" and "communist"; but it hasn't heard of "haddock" or "avocado". It knows Picasso and Gaugin, but seems to be unacquainted with Raphael and Leonardo. Likes Elvis and Dylan; hasn't heard of the Beatles. Seattle, Quebec and Tampere (Finnish city, pop 200,000), yes; Newcastle, Cardiff and Sheffield (English city, pop half a million), no.
If you expect your Nokia to know what you're on about when you message your partner, "That shop in Bolton has the grout we need," it will begin grumpily demanding, "Spell?" If, however, you want to say: "Who's your favourite dictator, Stalin, Hitler, Franco or Napoleon?" the mobile understands exactly what you mean. My phone goes all talk-to-the-hand if I try to message somebody to remember to get the emulsion, and that Sainsbury's has guavas. But write: "Marxist dogma relies too heavily on the dialectical approach," and the Nokia begins, figuratively speaking, to nod in agreement.
On the basis of its vocabulary, in short, my 21st-century Nokia phone, if it were a person, would be a heavily bearded lecturer from the London School of Economics in 1975, smelling strongly of pipe smoke. And I'd be the last person to want that academic out of my mobile. I just wonder whether he might not be joined in there by someone a bit more UK 2006-specific.
Nokia's text dictionary for the English-speaking world (there aren't separate ones for Britain and America), like that for most mobiles, is provided by a Seattle-based outfit called Tegic Communications, now owned by the AOL empire. The original database of words - it has grown to 34,000, including plurals and different forms of the same word, against more than 600,000 in the OED - came from a pre-mobile piece of software which speeded up communication for disabled people only able to tap one letter at a time.
Tegic won much publicity last month when it boasted about adding funky coinages such as "lifehack" and "podjack" to its predictive dictionary. I asked AOL's Erin Gifford whether there was any chance of getting perhaps less trendy, but more useful words, such as "Wolverhampton", "plaice" or "Heathrow". It turned out that Tegic's methodology was quite old-media.
"When we put the dictionary together, we pay attention to current newspaper data," Gifford said. "It's not the 34,000 words most used in texting, it's the words most commonly found in print. Heathrow and Sheffield aren't in there, but London and Edinburgh are. It just depends on how frequently they occur in the media. Avocados are not written about enough to have made it."
Gifford said Tegic might respond to a write-in campaign urging the inclusion of a particular word, but given the firm's usual method, it might be better to use the last bit of space I have left to add usage heft to the vocabulary of the rejected: avocado! Cardiff! Can you hear us in Seattle? Haddock! Mozart! Mozart!
