On the morning of April 1, I heard a BBC newsreader say (without levity, April Fool’s Day though it was) that Sajid Javid, the British government’s secretary of state for business, innovation, and skills, had “assured the steel workers that ministers were doing everything they could to save their jobs.” And for a few misguided milliseconds my brain was saying “Typical: politicians trying to protect themselves!” I had linked the genitive pronoun their to the most recent plural noun, ministers.
The scriptwriter intended something very different, of course. The Tata Steel megacorporation has been running British steel plants at a loss of more than $1.4 million, not per annum but per diem. When Tata announced last week that it planned to sell or close down this loss-making operation, government ministers knew it would mean the disappearance of thousands of jobs, and catastrophe for a town like Port Talbot, in Wales, that depends almost entirely on its steel plant. Mr. Javid faced instant criticism for not being on the case — a trifle unfair, since he was conducting government business 12 time zones away in Australia.
But he made the grueling 24-hour flight back, and went straight to the Port Talbot steelworks to be shouted at. What he told them was that the government was considering all options for keeping the plant open to stop workers being laid off. We were supposed to associate their in the BBC sentence with the more distant plural noun, workers.
This sort of antecedent-resolution problem with pronouns is ubiquitous in both speech and writing. The notion that properly prepared prose avoids ambiguity is an absurd myth. Repeating it to freshman-comp students won’t make it true. If ambiguity were even mildly toxic, we’d all be dead. It surrounds us constantly.
When linguists discuss semantics they employ subscript variable letters (often i, j, k, ...), somewhat analogous to the variables in logical calculi and computer programming languages. So they might write [1] for the meaning where ministers are engaged in protecting their own positions, or [2] for the one where they are helping the workers.
| [1] | Steel workersi should understand that ministersj are doing everything theyj can to save theirj jobs. |
| [2] | Steel workersi should understand that ministersj are doing everything theyj can to save theiri jobs. |
It’s a very interesting fact that no spoken language has ever been found to include in its vocabulary an extensible stock of markers to signal where the linkages are between pronouns and their antecedents.
Some languages, like Navajo (Diné Bizaad), have a distinction between third person and fourth person (though in Navajo it is expressed in verb inflections, not pronouns of different shapes). The meaning difference is subtle; some sources say that definitely identified individuals get third person and vaguely delineated or inferred individuals get fourth person. Apparently when talking to others about a man who happens to be present, it is courteous to refer to him with fourth person, which sort of keeps the reference at arm’s length. I’m unqualified to give further details (I’m not an Athabaskan specialist; as the late great linguist James McCawley would sometimes say after a class exposition of some esoteric point, “I’ve already told you more than I know on this topic”).
What I do know, however, is that although grammatical features like person, gender, number, and case sometimes help to prevent confusion, no language has an open set of words or affixes used simply to prevent multiple references to distinct people from being confused where grammatical features don’t suffice.
Languages with such devices are easy to envisage. Instead of he, she, it, etc., English could have had multiple pronouns like heej, heek, heem, etc., and sheej, sheek, sheem, etc., and iteej, iteek, iteem, etc.; and matching noun suffixes -eej, -eek, -eem, etc.
Even a small number of them could helpfully clarify (for example) jokes about ethnically diverse sets of men walking into bars and suchlike. Here’s a sample joke in an imagined dialect of English. I’ve set the special suffixes in red for ease of reading. The Englishman gets -eej, the Scotsman gets -eek, the baker gets -eeb, the Englishman’s pocket gets -eep, etc.
An Englishman-eej and a Scotsman-eek walk into a fancy bakery shop. Heej snatches two cookies-eet off the counter and pops themeet into hiseej pocket-eep without the baker-eeb noticing.
“See?” heej mutters to himeek; “You’re just not smart enough.”
“Wrong,” heek replies. “Watch and learn.”
“I’ll show you a really good magic trick,” heek says to heeb.
Heek calmly picks up a cookie-eef and eats iteef. Then heek takes another one-eeg and eats iteeg too.
By this time, heeb is getting suspicious, and asks angrily what’s so damn magical.
So heek says to heeb: “Now look in heej’s pocket-eep!”
All pronoun ambiguity gone. It would be as good as having subscripts or variable letters. But in all of the world’s 7,000-odd spoken languages there isn’t one that attempts anything like this, despite how much it might help comedians, newsreaders, lawyers, or the rest of us.
