It is impossible for an Englishman to open his mouth without making some other Englishman despise him.

George Bernard Shaw, introduction to Pygmalion (1913)

Alan Ross was Professor of Linguistics at Birmingham University when he published this paper in Neuphilologische Mitteilungen, the journal of the Modern Language Society of Helsinki, although The Dictionary of National Biography tells us he started writing it as “an undergraduate amusement at Oxford”. So it had simmered on his back-burner for a quarter-century or so before it finally saw the light of day. (One can imagine Ross, now academically senior and secure, deciding it would be a bit of a lark, and could now do him no harm, to place his “undergraduate amusement” in a learned journal.)

This is the paper that gave the world (or at least a select English-speaking part of it) the idea of “U” and “non-U” usage—patterns in speech and writing which (Ross suggested) distinguish the language of the English upper class from that of the aspiring middle class.

In this article I use the terms upper class (abbreviated : U), correct, proper, legitimate, appropriate (sometimes also possible) and similar expressions (including some containing the word should) to designate usages of the upper class; their antonyms (non-U, incorrect, not proper, not legitimate, etc.) to designate usages which are not upper class.

The paper would probably have fallen into obscurity had it not been read by Nancy Mitford, who referred to it in an article entitled “The English Aristocracy”, published in Encounter magazine (1955). This was sufficiently well-received that Mitford went on to edit Noblesse Oblige: An Enquiry Into the Identifiable Characteristics of the English Aristocracy (1956), which contained her original essay and a condensed version of Ross’s paper, together with contributions from Evelyn Waugh, Peter Fleming, Christopher Sykes and a poem (“How To Get On In Society”) by John Betjeman.

And so Ross’s distinction between “U” and “non-U” became the 1950s version of an internet meme—it was a topic for excited and amusing conversation, to which people could make their own contributions, and which gave those in the know a sense of “in-crowd” satisfaction. This led to Ross’s original paper becoming “famous by repute but paradoxically little-known”, to quote from Clive Upton’s introductory comments, written when the paper was reprinted by Neuphilologische Mitteilungen in 2007.

So I thought I’d take a look at it.

Ross gives his reason for studying the difference between U and non-U usage as follows:

It is solely by its language that the [English] upper class is clearly marked off from the others. In times past (e. g. in the Victorian and Edwardian periods) this was not the case. But, to-day, a member of the upper class is, for instance, not necessarily better educated, cleaner or richer than someone not of this class.

He then admits that there are some minor differences in manners and interests between the upper class and the others—his list includes “an aversion to high tea” and “not playing tennis in braces”. It’s at this point one begins to suspect that Ross may not be entirely serious. This is confirmed when he goes on:

Again, when drunk, gentlemen often become amorous or maudlin or vomit in public, but they never become truculent.

Truculence, I deduce, would be the mark of a cad. Or perhaps a bounder.

Ross presents his linguistic analysis in two sections. The first is “The Written Language”, which deals with how people address envelopes and postcards, and the forms of salutations and letter-endings. This really does little to contrast U and non-U usages—it’s merely a list of ways of doing things that are considered “correct”, easy mastery of which Ross identifies as a U domain. So, for instance:

Modes of address, particularly those used for the nobility, have always been a bugbear to the non-U. It is, for instance, non-U to speak of an earl as The Earl of P—; he should be spoken of and to as Lord P— and also so addressed at the beginning of a letter if an introduction between him and speaker/writer has been effected.

Many of these rules of etiquette in writing are still around; some are now gone, but familiar to me from my schooldays, like:

Letter-endings. The U rules for ending letters are very strict; failure to observe them usually implies non-U-ness, sometimes only youth. In general, the endings of letters are conditioned by their beginnings. Thus a beginning (Dear) Sir requires the ending Yours faithfully […]

Well, duh. Miss Macpherson, my fearsome primary-school teacher, got that one well drummed into this (very non-U) youth before I turned twelve.

The section ends with a dissection of the advice given by R.W. Chapman in Names, Designations and Appellations (1936*), published by the Society for Pure English, no less. In general, Ross thinks that Chapman’s reflections were dated, even for their time of writing. To Chapman’s discussion of the use of surnames, Ross adds an observation about U schoolchildren addressing each other by their surnames only (as in the “Jennings and Darbishire” books by Anthony Buckeridge):

It is not until a boy gets older (c. 16?) that he realises that he must deliberately ascertain his friends’ Christian names in order to be able to refer to them correctly to their parents. At Oxford in the late twenties the use of the surname in these circumstances was a known gaucherie and must therefore have been fairly usual.

I also rather relish Ross’s personal anecdote on the use of the suffix “Esq.” (“Esquire”), which in Ross’s day was a standard way of addressing a letter to a gentleman.

Knowledge of at least one initial of the recipient’s name is, of course, a prerequisite for addressing him with Esq. If the writer has not this minimum knowledge (and cannot, or is too lazy to obtain it) he will be in a quandary. In these circumstances I myself use the Greek letter θ (as θ. Smith, Esq.) but this is probably idiosyncratic.

I think it may well be, Professor Ross.

Section Two, “The Spoken Language”, begins with pronunciation. This section is a bit difficult to follow for a modern reader, because Ross’s phonetic notation predates the standard International Phonetic Alphabet. But, among other things, we learn that U-speakers say “temprilly” rather than “temporarily”; drop the letter l from “golf” and “Ralph” (“goff” and “Rafe”); and pronounce “tyre” and “tar” the same way (“taa”). “Either” pronounced with the first syllable sounding like “eye” is U; first syllable “ee” is non-U.

Ross then asks, rhetorically, if it is possible for a non-U speaker to become a U-speaker. He concludes that there is only one way:

This is to send him first to a preparatory school, then to a good boarding school. This is a method that has been approved for more than a century and, at the moment, it is almost completely effective.

Good to know.

It’s not until we get to Part 6 of Section 2 that we reach the bit for which Ross is now remembered, and around which most of the U/non-U excitement took place in the 1950s—U and non-U vocabulary and phrasing, arranged in an itemized list. So we’re told that “They’ve a lovely home” is non-U; “They’ve a very nice house” is U. Non-U people call the lowest-ranking face-card the “jack”; the U equivalent is “knave”. “Serviette” is non-U; “table-napkin” is U. And then there’s this:

Coach. ‘char-a-banc’ is non-U, doubtless because the thing itself is. Those U-speakers who are forced, by penury, to use them call them buses, thereby causing great confusion (a coach runs into the country, a bus within the town).

Finally, and to me unexpectedly, Ross addresses the issue of ephemerality—the fact that what’s U for speakers in the 1950s was not so in the 1850s or 1750s. To make his point, he quotes extensively from relevant sections of John Walker’s Critical Pronouncing Dictionary and Expositor of the English Language (1791). His conclusion is:

The ephemeral nature of our present system of linguistic class-indicators is very clear from the above citations from Walker. Nearly all the points mentioned by him—only one hundred and sixty years ago—are now “dead” and without class-significance, in that one of the pronunciations given is to-day no longer known in any kind of English save dialect. […] In three cases of double pronunciations, to-day’s U alternative is chosen by Walker as the non-U one.

Which is refreshing—Ross is too much of a linguist to believe “that the customs of his tribe and island are the laws of nature” (to quote George Bernard Shaw again).

So this is, in essence, a glimpse of the language spoken by the English aristocracy in the mid-twentieth century, and the shibboleths they used to identify an out-group by which they undoubtedly felt somewhat threatened. And Ross is certainly an entertaining guide.

I haven’t been able to find a version of Ross’s original paper that doesn’t require either an academic institution log-in, or signing up to some document server service. You can, however, read the condensed version that appeared in Noblesse Oblige here. He went on to write a series of popular books on English usage and pronunciation: What Are U? (1969), How To Pronounce It (1970) and Don’t Say It! (1973). I haven’t read any of these, but if I get around to it, I’ll report back.

---

* Ross’s article misprints this date as “1946”