The Incompleat Gamester
Nav to sitemap


On the acceptability of words in word games
by David Parlett

This is an expanded and updated (2013) version of an essay that first appeared in the Penguin Book of Word Games, published in America as Botticelli and Beyond (1982) and as THE Book of Word Games by Echo Point Books (2012). (Not my choice of title but I refused to allow it to be called The Ultimate Book of Word Games .) It is written in, and from the viewpoint of, British English. References to the Oxford English Dictionary are to the constantly updated online edition. (See how to subscribe if you don't already.) An invaluable on-line resource for all serious word-gamers is Michael Quinion's World Wide Words.

Note: I've removed all the links to Ogden's Basic English from this page because the sites they led to were insecure. Instead, see the Wikipedia article

Word games differ from most other games in a distinctive double feature which can be either charming or exasperating, depending on which barrel is loaded and which end of it you happen to be looking down at the time. This is the question whether a given sequence of letters does or does not exist as a word, and, if so, whether it should be acceptable as such in a given word game.

If you were learning Poker, you would be told that a full house is a hand containing three cards of one rank plus two of another. Thereafter, if dealt (say) three Sixes and two Kings, you would be in no doubt that you held a full house. If you were an alien learning English word games, you might be told that a word can begin with CH, that CH can be followed by EA, and that EAL is a possible word ending. Thereafter, confronted with CHEAL - you would not be any the wiser as to whether it were a word or not. CHEAL is only a word if it has a meaning. WHEAL has a meaning; so does CHEAT; so even does WHEAT. But CHEAL?

Any three cards of one rank and two of another make a full house, but the combination of any three consonants and two vowels makes a word only if it happens to be assigned a "meaning" in an alphabetically ordered list of otherwise quite haphazard letter-combinations known as a dictionary. It is this asymmetry or unpredictability of the material that makes word games so appealing (or appalling, if you're at the wrong end of the barrel). Words defy logic. So do most people. Hence the popularity of word games.

But it is an irregularity that can hinder the enjoyment of word games by causing arguments. How can this problem be overcome?

One way is to agree beforehand that words only count if they are listed in the nearest dictionary to hand. This may be called the Rule of Authority. Even as it stands, it causes complications due to the fact that dictionaries not only differ from one another as to which words they have the space or inclination to recognize but also tend to be inconsistent within their own covers. For instance, Chambers Dictionary, widely used by British word-gamers, lists some interjections but not others, recognizes the verb cuddle but not the adjective cuddly, and writes Swede (the vegetable) as if it were a personal name. In any case, word-gamers do not normally restrict themselves to dictionary words but go further by refusing to recognize certain categories, such as unnaturalised foreign words, words spelt with a capital initial, and the verbal forms of letters of the alphabet.

Thus the Rule of Authority is bad in principle, because it wrongly assumes that the language is fixed and immutable, and bad in practice, if only because it is abused.

In its place, therefore, I wish to propose the Rule of Appeal. This does not do away with dictionaries, but relegates them to a position of last resort. It throws the responsibility of word recognition on to the players themselves. To be effective, it requires players to make disinterested judgements which may conflict with their own positions at the time of appeal. This may cause problems among players with an overdeveloped sense of competitiveness, and therefore cannot be recommended to their attention: they will be better off with the more dogmatic Rule of Authority. It has always proved effective in my circle of players. It runs as follows...

uplink uplink

The rule of appeal

A word is acceptable if it

(a) exists,
(b) is spelt correctly, and
(c) accords with the conventions of the game.

If a word is challenged under any of these heads the claimant may appeal to the other players. If at least one opponent accepts it, then it is acceptable. If not, the claimant must either withdraw it or win support by using a dictionary or other source to demonstrate the word's existence, spelling or conformity with agreed conventions.

Obviously, in a two-player game there is likely to be more emphasis on recourse to printed references than when more are taking part. However, the principle should still be as far as possible that language was made for man and not man for language: dictionaries can only follow and lag behind the usage of speakers, to which they should be regarded as standing in subservience rather than authority.

In applying the Rule of Appeal, a dictionary can be used to demonstrate that a word "exists". The dictionary may list it as a main or subsidiary entry or draw attention to it in some other way. If not flatly contradicted by the dictionary, the word may be inferable from a main entry by application of the normal rules of English word-construction. For example, nouns are pluralised by the addition of S unless otherwise stated - so, for example, the listing of children as the plural of child negates the form *childs. (The star denotes a non-existent or unattested word.) But even here dictionaries will let you down. From Chambers, for instance, you can infer ifs and buts but not whys and wherefores.

Words may be omitted from dictionaries for many reasons. The simplest is that there are so many words that only the largest could seek to list them all: the shorter they are, the more they must omit. So-called source dictionaries, such as the Oxford English, support their material by quotations from literature in which the various words and their forms have been used. Such works take years to compile and are out of date by the time they appear. Newer dictionaries may make some use of real sources but seek primarily to update existing material ultimately derived from source dictionaries.

In these circumstances it seems reasonable to allow players to support claimed words by reference to primary sources. In other words, if you can demonstrate that the word or form of word which you are claiming has been used in print in an intelligible way, and without attention being drawn to it as a coinage, that should be sufficient to persuade an opponent to deem it acceptable. This resource should be used with discretion - or, at least, with a time limit; otherwise the burden of proof could hold the game up by going on all night.

Another way of seeking support for a claimed word is to propose a fluent sentence which contains it naturally and intelligibly and without drawing attention to itself as a coinage. This applies especially to words formed by the addition of affixes such as re-, un-, -ment, -isation, etc. I would not expect any dictionary to list, for example, unvowed, but cannot find any objection to its use in a sentence such as "The candidate left no electoral promise unvowed".

Similar remarks apply to the question of spelling, but with provisos. A general proviso is that players should agree beforehand whether to accept both British and American alternative spellings or only those applicable to themselves. On the whole I favour the latter, though it could give rise to problems in mixed British and American company. Other alternative spellings may be accepted if listed in the dictionary. Otherwise, however, it is not wise to go to primary sources. Many a journalist commits such spelling errors as *supercede, *forsee, *ancilliary and so on, and even books have been known to contain mirspints. As for proposing a sentence containing your own spelling... well, I spoze you could try it on, but don't expect it to be accepted. (Though it's been more likely to occur since the advent of texting and, more horrendously, Facebook; but that way madness lies.) The spelling of words is more open to authority than their existence. Spellings are conventional rather than natural, and their maintenance is a matter of practical desirability rather than elitist tyranny. Written communication would break down if we all spelt as the fancy took us.

The third part of my Rule of Appeal - accordance with the conventions of acceptability - recognizes the fact that most players agree to count certain types of words as unacceptable, though different circles rarely agree as to which categories they ban. One extremist school of thought maintains that word games should test only one's active vocabulary, i.e. those words which players themselves are likely to have used in context over the past year. In my case this would cut out such unexceptional words as jeopardy and coax, which I am quite convinced I have not used in the past ten years, if ever. (That is to say, ten years ago it occurred to me to wonder if I had ever used them; since then I have been keeping a careful watch but have never found myself in a context where either word would be the most natural choice.) Literate people's passive vocabulary contains many more words that they recognize than they are likely to use. Even so, there are certain words which seem to have any active existence only in word games - things like al and jo and zax - and I for one would be quite happy to ban them if it were at all possible to draw a sensible dividing line between a vocabulary active only in games and a passively communicative one. The only practical way of doing this, however, would be to restrict play to the use of no more than the 850 words of Ogden's "Basic English" vocabulary, which was promoted during the 1930s as an international auxiliary language. But this would probably not make for a wildly exciting game of any sort.

I share most players' desire to avoid words which exist only as proper names and have some sympathy with their dislike of not-yet-naturalised foreign words. I cannot understand the objection to "slang", or to the verbal forms of letters of the alphabet - aitch and ar, to name two - but dare say that many would not share my aversion to interjections in general and those of two letters, such as fy and st, in particular. (If you're having st, I want zz.) Indeed, I would do away with two-letter words altogether: some games would improve immeasurably as a result.

The trouble with any proposed rule to allow or disallow certain types of word is that lines have to be drawn somewhere, but words do not fit into logically circumscribed categories capable of such delineation. If rules are interpreted by the spirit, opposing players will disagree because of human nature. If they are interpreted by the letter, such absurdities occur as a once reported ruling of the British National Scrabble Championships permitting the plural forms of interjections. Ahs may just about pass muster ("This prospect was greeted with many oohs and ahs"), but, really, would you seriously allow shs?

It may be amusing, even instructive, to see what other problems and absurdities can arise from the various criteria of acceptability open to adoption by word-gamers.

uplink uplink


"They are the most economical of all words", declares Sir Alan Gardiner in The Theory of Proper Names, "inasmuch as they make only a very small demand on the eloquence of the speaker, and an equally small demand upon the attention of the listener."

A clever remark, and, like most clever remarks, highly debatable. If I were listening to a speech in which I was enjoined to show the qualities of a Codrus I should feel that a very great demand was being made upon my attention, and obliged forthwith to consult my copy of Betty Radice's invaluable Who's Who in the Ancient World. Of course, I wrote that bit before the days of the Internet.) Gardiner was evidently no word-gamer, otherwise he would have realised that it is precisely because they make too great a demand on the attention that most players refuse to accept them as genuine word-game fodder.

In theory, the difference between a proper name and a common one is that the former refers to a set containing only one member, whereas the latter is a generic term denoting a whole category of similar members. The motive of word-gamers in rejecting names is that acquaintance with them is a test of general knowledge, whereas word games are intended to be a test of literacy. It is reasonable to expect word-gamers to have mountain in their vocabulary; it is not reasonable to expect them to know Matterhorn any more than Makalu or Mercedario.

In practice, of course, many names do not have unique referends. I am not the only David in the street, let alone the world. There could be as many Davids as there are dogs, and it is reasonable to expect literate people to have both words in their vocabulary. Even so, a name is felt to be unique rather than generic, as may be gauged by the incongruous effect produced by ssuming the opposite in the following encounter from The Lost Weekend (Paramount, 1945):

Male nurse - Name?
Ray Milland - Birnam.
Male nurse - What kind of a Birnam?
Milland - DON Birnam.

Proper names are shared only because there are not enough of them to go around. Dog is generic because it will do for any individual member of a genus of animals which we are all capable of recognizing, but we could not look at a person and know for certain that it was a Birnam or a David.

At first sight, English appears to help word-gamers by its convention of writing names with capital initials. Even more advantageously, names which become used as generics drop their capitals. Thus moll, originally a proper name (diminutive of Mary), comes to mean a gangster's girl-friend, in which capacity it takes a small M and is acceptable as a generic word or common noun. Similarly, quisling is now acceptable as a common-noun synonym for traitor and boycott is a verb. (More surprisingly, guy is from Fawkes and gun from Gunhilda.)

Unfortunately, this is not always the case. We may refer to a poet as a Shakespeare or a Byron, but not with a small S or B. It all comes down to a question of whether or not we feel certain names to be "proper" in any given circumstance. Once again, we trip over the line that can't be drawn.

More seriously, because unnoticed by many players, English happens to use capital initials for words whose status as "proper" nouns is highly questionable. What about days of the week and months of the year? I regard Sunday, January, Monday, February and the rest as generics and favour the French custom of spelling them with small initials. Yet so deeply ingrained is the contrary behaviour of the English that I have even been challenged for (quite correctly) spelling the four seasons with small letters instead of capitals!

Again, what about nationals? The French, bless their logical minds, quite properly spell their words for Frenchman, lcelander and so on with a small initial. It seems obvious to me that Frenchmen are generic rather than proper. The same applies to languages, though with less force: I can see the argument behind treating French as a proper name, in so far as it denotes a unique language. Less defensible is our habit of spelling adjectives of nationality with a capital initial, except in given phrases. What is the difference between a French widow and a french window (apart from the shape)? Not just the N in window, but the F in French as well.

Many word-gamers also disallow the so-called "names" of letters of the alphabet. To my mind, one aitch is as generic as another. Nobody insists that they be spelt with capital initials, but that doesn't stop them from banning the words. And if names of English letters are prohibited, what about Greek and Hebrew? If pi is acceptable because, besides being a Greek letter, it denotes a mathematical quantity, isn't the Hebrew aleph precisely analogous to the extent that it denotes an order of infinity? Are el and es more "proper" than six and seven?

Problems of consistency can arise with names of plants and animals. All creatures great and small have scientific names, great and small, taken mainly from Latin. But whereas no one would object to wolf in a word game, neither part of its scientific name, Canis lupus, would be accepted, the former if only because it is (by international convention) spelt with a capital, the latter if only because it is Latin and therefore "foreign". There is no problem here because we have a so-called popular word (wolf), which is quite distinct from the scientific. But problems can arise where no popular word exists. Dinosaur has long been a popular word and was never very scientific in the first place, but word-gamers may quite naturally balk at Stegosaurus or Triceratops. Strictly speaking, or writing, they are spelt with capitals; whether or not they should be so spelt in popular writings is an open question. I pose the question only to be awkward, not to answer it, though I dare say most players would agree with me in accepting them on an equal footing with dog and wolf.

The gist of my argument - in case it has got lost in the undergrowth - is that proper names should not be allowed in word games because they belong to specialised knowledge rather than to literacy. The task of deciding which names are proper and which generic, or common, is too elaborate for practical play. It is therefore a poor best to prohibit words which in English are only spelt with a capital initial, even if this means throwing out with the specific bathwater such generic babies as Monday, Zed and Tyrannosaurus rex.

I leave this subject with a few names of ponderable propriety. What about Londoner? Mancunian (pertaining to Manchester)? Petrine (epistles)? Caesar? Mackintosh? Chevvy? Formica? Spa? Lido? Thermos?

And did you know that the plural of thermos is thermoi?

uplink uplink

Names as generics

Many words which may be disallowed from word games because they are chiefly known as names have, nevertheless, a generic meaning or usage for which a capital initial is not required. Following is a list of personal names which are also acceptable with a small letter as game words. They are chiefly, but not exclusively, personal forenames. Others are family or place names. I make no claim to completeness, and in any case have deliberately omitted are names which are too well known as common words to need listing, such as floral names like pansy, heather, violet, etc.

abigail, abram (obs.), alexanders (plural only), alexandrine, alma, amazon, amelia, ann (scot.), anna, ava, barney, basil, benedict, benjamin, bertha, beryl, betty, biddy, billy, blanch, bob(by), carol (but not Carole), cecils, (plural only) celeste, celestine, charlie, -ey, charlotte, christiana, christie, -ey, cicely, clarence, clement(ine), (without -e), collie, craig (Scot.), crispin, davenport, derrick, dick, -y, -ey, dickens, ebenezer, emma, eric (but not Erica), fanny, fay, flora, flossy (but not -ie), french, gemma, gene, georgette, gloria (Latin), gregory, harry, henry, homer, ingram (obs.), iris, isabel(-la) (but not Isobel(la), jack, james, jane, jean, japan, jasmine, jeff, jemima, jemmy, jenny, jereboam, jerry, jersey, jess, jill, jimmy (US, not UK), jo, joanna, jobe, joey, jo(h)annes, john (US, not UK), joss, jud, judas, judy, julienne, juvenal, kelvin, ken, kitty, lambert, laura, lewis, lill (Scot.), lionel, louis, madge, mag, madras, magdalen(e), margarite, marguerite, martin, maud, mavis, maxwell, may, mick(-ey, -y), mike, mina, moll(y), mona, mungo, myrtle, nancy, nanny, neddy, nelly, ngaio, norma, norman, oliver, otto, paddy, pam, patty, paul, peggy, peregrine, perry, peter, poll(y), regina, rex, reynard, roger, rory, rosemary, ruth, sally, saul (Scot.), sheila, sophia (but not Sophie), stephane, stepney, talbot, timon (obs.), timothy, toby, tom(my), tony (obs. slang), valentine, valerian, veronica, vesta, victor (but not Victoria), wally, wellington, willy.

I compiled this list by checking through Chambers headwords and concentrating on personal forenames. The name Constance is proper only, though the 0ED records it as obsolete for "constancy". Chambers does not record jenny (of spinning fame), joanna (for "piano"), rex without a capital initial - all of which I would accept - and is ambiguous about roger. However, the inadequacy of Chambers in this respect is demonstrated by its listing of swede only with a capital initial, even though no literate person would so spell the vegetable of that name. Some names, such as Vanessa (denoting the Red Admiral butterfly), cannot be spelt with a small initial even when not used personally, since, by international agreement, a genus name is always capitalised. This should not, however, debar names of genera for which the scientific and popular words are identical, such as pteranodon.

Then, of course, there are place-names also used generically, such as chesterfield, ulster, and many, many more.

uplink uplink

Foreign words

Talking of thermoses reminds me that of all the restrictions commonly followed in word games none is more laudable than the prohibition of foreign words, smelling as they do of goat-skins, garlic, and cheap red wine. English speakers rightly take it for granted that the game is to be played in English: if foreign words are allowed an unfair advantage accrues to the most accomplished linguist among them.

Unfortunately, though laudable, the rule is unenforceable. There is a line, again, that won't be drawn.

The problem arises because English has a habit of picking up foreign words and gradually absorbing them into its lexicon until a point is reached at which they cease to sound foreign. They are then not foreign but only of foreign origin. By this process, more English words actually turn out to be foreign than native: it has often been pointed out that English is a Germanic language with a vocabulary of which some seventy-five per cent is Romance (ultimately derived from Latin).

It is aggravated by the fact that no two dictionaries, and often no two players, will agree on the extent to which a word in question has becom naturalised. The Oxford English Dictionary attaches a sign to words not regarded as naturalised, but as it was compiled over a century ago this is not very helpful. (Though it has been updated since I wrote this essay.)

English easily absorbs foreign words for both phonetic and syntactic reasons. It has a large, rich sound system which can accommodate almost any foreign word with a fair approximation to its original sound - although, in practice, English speakers have a notorious tendency to Anglicise beyond recognition. An interesting example of phonetic flexibility is the English speaker's readiness to reproduce passably the sound of ch in loch or gh in Van Gogh, even though the sound dropped out of English itself several centuries ago. And the fact that English makes minimal use of grammatical modifications such as case-endings means that the absorption of foreign words is not hindered by problems of declension or conjugation.

Being, therefore, in a position to employ foreign words more or less indiscriminately, English readily does so in several different ways:

  1. Unlike the Romans, whose pains to translate everything into Latin extended even to the translation of Celtic gods into Mars, Apollo and so on, English-speakers quite happily retain native words for things peculiar to their country of origin. Thus, for the price of one zloty you may drink slivovitz while playing the crwth in a troika. Olé!
  2. English-speakers will often not trouble to find equivalents for words denoting things of foreign invention, albeit of international applicability. Hence, the pen of my ombudsman is in the bidet.
  3. Even where an English synonym exists, the native word may be retained for greater accuracy or local colour. Both pueblo and kampong can be replaced by village, but this does not prevent them from being listed in Chambers, the dictionary most favoured by British word-gamers. An interesting tailpiece to this observation is the use of the native spelling cwm for such a valley as may be found in Wales, from where the name was borrowed and Anglicised to coomb in the first place.
  4. Foreign words are often retained for abstractions where, again, the use of English equivalents may lose in the translation. Apartheid was always likelier to lead to angst than to détente.
  5. If the three words just quoted are used as much for emotional colouring as for accuracy, others are employed for accuracy without the slightest taint of emotive connotation. Such are mainly technical terms, largely from Greek and Latin, with examples rife in biology and medicine: rubella, sinus, the genus Homo. It is interesting to note that Chambers lists manus for hand but not pes for foot, though both may be encountered in the same context. To this category belong also Italian musical terms such as allegro, crescendo, and niente.

All the foreign words quoted above are, to my mind, acceptable as English in word games, and I should like to think that most players would agree. At the same time, I cannot believe that any self-respecting word-gamer would accept the German Gesellschaft for company (commercial), even though it is listed in Chambers.

On the whole, I have a fairly liberal attitude towards foreign words. I tend to accept a word unless the dictionary specifically describes it as foreign or ascribes it to a foreign language other than to indicate its origin. Doubtful cases may be tested by asking the claimant word-maker to propose a sentence or context in which the foreign word would be the natural choice of an educated English-speaker.

As a postscript, I do object to the use of individual foreign words abstracted from phrases or sayings. Neither desperandum from nil desperandum, nor cul from cul-de-sac, nor sturm from Sturm und Drang, nor istesso from l'istesso tempo, nor hoi from hoi polloi has independent existence as a word in English contexts. All should be disallowed for that reason.

uplink uplink

Dialect words

Most word-gamers quite properly do not count dialectal variants of English as "foreign" and consequently allow dialectical words as valid. But few go so far as to discuss the point, which I think is a little short-sighted as there are various pitfalls which can lead to argument. The matter is not helped by the fact that, although everyone knows more or less what they mean by "dialect" there is no precise linguistic definition of it. The term is often employed patronisingly out of ignorance. I recently read that a WANTED FOR MURDER poster circulated in England appeared not only in English but also in the Indian "dialects" of Urdu, Punjabi and Gujarati. Whoever wrote this evidently assumed that different languages, if spoken in the same state, are dialects of one. Pursuit of this argument would lead to the preposterous classification of Welsh and Gaelic as dialects of English.

Dialect is not a strong term amongst the English because regional varieties of the notional "standard" vary little from one another compared with those of more strongly differentiated languages such as Italian. The only English dialect sufficiently well marked to boast literary status is Lallans or Lowlands (Scots). (It is sufficiently marked to be discussable as a language. A Dutch friend of mine once reported that he switched on the radio one Sunday morning to hear a church service conducted in what he thought was Frisian. It turned out to be Lallans.) A goodly proportion of the words in many English dictionaries are designated "Scots", and, although I do not necessarily disagree with their acceptance, I am surprised not to have read any questioning of their acceptability in books and articles on English word-gaming. Should Sassenach word-gamers be expected to know such words as kye for cows, fley for frighten, sark for shirt - for any purpose, that is, other than to show them off in word games? And if these, which appear in Chambers, what about those that do not: mebbies for perhaps, mower (a compound preposition), throuither for jumbled up?

British word-gamers south of the border may be forgiven their greater knowledge of American English than of Scots. The general meaninglessness of the word "dialect" may be indicated by noting that, to Britishers, American is a dialect as opposed to a standard, while, to Americans, British is equally a dialect. It is interesting to note that British players will generally accept American words as valid (movie, debunk, racketeer, etc.) but not American spellings of words common to both languages (harbor, kilometer, defense). At least, literate people such as word-gamers do not. Professional illiterates, such as journalists, sociologists, and information technologists, tend to use American spelling as much as American vocabulary, probably without noticing it.

In the rules of the British National Scrabble Championship that I consulted when originally writing this essay I noticed that words listed Chambers only as "U.S.", such as bronco, were disallowed as being foreign, and found this extraordinary, in view of the unquestioning acceptance of vocabulary listed only as "Scots". Since then, however, Chambers has published OSW - Official Scrabble Words, which lists everything together indiscriminately, including - what strikes me as equally ridiculous - both British and American spellings of the same words.

(By way of update, I've just been checking the online version of OSW and am interested to see that it includes fley but not throuither.)

uplink uplink


It is a commonplace of knowledge that languages evolve, losing and acquiring new vocabulary and undergoing shifts and drifts of grammatical structure. We have already commented on the gradual absorption of words which start out as foreign and gradually become so naturalised that the discovery of their origin comes as a surprise.

In counterpoint to the absorption of new vocabulary runs the gradual shedding of the old. When a word ceases to be current it has started on the road of obsolescence: it becomes first quaint, then old-fashioned, then archaic, and finally obsolete. These are not precise terms, of course; the process is more gradual than their use implies.

The general practice of word-gamers is to prohibit words which have reached the final stage and are accordingly marked "obs." in the dictionary. As to archaic words, such as thee and thou, surviving only in specialised contexts, lips may be pursed and acceptance grudgingly granted. Fair enough; but then problems arise with the appropriate verb forms. If thou is permitted, so must be hast; if hast, why not also hath, seeth and wert? And if these are acceptable because they can easily be demonstrated in print, what about the anachronistic [thou] faxest? Where do you draw the line?

Care must be taken not to dismiss words as obsolete or archaic merely because the thing they refer to has vanished from the scene. Groats are obsolete, but groat is not. The word remains current to the extent that it has not been replaced by a modern word for the same coin: you may yet dig one up in the garden and will know exactly what to call it. Granted, you are more likely to unearth a sixpence or a shilling, but the same principle applies. In similar vein, mermaids and unicorns may not exist, but mermaid and unicorn undoubtedly do.

The same might be argued of words listed only as "Shakespearian" or, worse still, "Spenserian" - but not by me. Personally, I reject them as obsolete. This may be challenged as inconsistent with the argument supporting groats and unicorns; but was it not somebody-or-other who said "Consistency is the bugaboo of petty minds"? (Yes - since writing this essay I have learnt it was Ralph Waldo Emerson.)

Obsolete spellings are more easily dealt with - i.e. rejected - on the ground that no literate person would use them in preference to the modern spelling. It should be noted, however, that some words have equally common variant spellings, such as (window-)sill and cill. If the main entry in a dictionary quotes alternative spellings they are generally accepted by word-gamers, and there seems nothing wrong in that.

uplink uplink


My original essay failed to mention the category of neologisms and coinages - words "invented" for particular purposes at particular times. They have always existed, but seem to have become especially prominent in the past quarter century or so, especially for sociological cohorts - perhaps initiated by the now somewhat dated yuppie for the "young, upwardly mobile" elements of society. What brought this category to mind was the fact that one of my circle of players recently argued against accepting prequel, on the grounds that it wasn't a "real" word, being of illicit etymology and in any case unnecessary since its function was perfectly well served by prelude. But this does not in fact serve quite the same purpose. A prelude is certainly the first part of a story, but it is integral to the work containing the story, whereas a prequel is, as everyone understands and the OED so defines, "A book, film, etc., the events portrayed in which or the concerns of which precede those of an existing completed work". To my surprise, it first appeared as early as 1973, and was used by as literary a figure as J R R Tolkien. Other invented words of so-called illicit etymology include television, which is a hybrid of Greek and Latin, and beefburger, consciously derived from a deliberate misinterpretation of hamburger - as if it were made of ham, rather than having originated in Hamburg. You may be surprised to discover that such invented words as transmogrify, quiz, and thingummy (spelt thing-o'-me by Fanny Burney) date from the 18th century, though the OED casts doubt on the story crediting the invention of quiz to a Dublin theatre manager called Daly. Even discombobulate goes back to 1834.

I'm inclined to say that if an invented word is in common use, and everyone knows what it means, there seems little point in rejecting it on grounds of either redundancy or linguistic purity. But, as always, you can take liberalism too far. I would certainly not accept such a barbarism as the all-too-often encountered miniscule, which is both redundant and ill-formed. Unlike prequel, which was coined to fill a lexical gap created by development of a new literary genre, miniscule is entirely redundant. We already have a perfectly good series of words denoting increasing smallness (small - tiny - minute - infinitesimal), leaving no gap needing to be filled. It is also ill-formed, being an ignorant cross between minuscule (meaning lower-case script or print) and minute, which I dare say many people shun because they misinterpret it as a unit of time, despite its distinctive pronunciation. To accept miniscule is to condone ignorance and laziness. Its opposites, the equally invented words ginormous and humongous, at least have the merit of deliberate jocularity, and are so recognized by the OED. (Surprisingly, OED records miniscule from as early as 1898, but still - quite rightly - labels it "erroneous".)

uplink uplink

Slang and colloquialisms

Some word-gamers prohibit slang, presumably on grounds of propriety. I don't object to current slang as such, but, as a parent (when I first wrote this, and now as a grandparent), I am in sympathy with the view that children should not be exposed to foul language in the home. It is true that the foulness of some slang lies not in the words themselves but in the attitude of the speaker who uses them; but, again, it is an attitude of aggression and lack of self-control which children should not be allowed to emulate. Adult word-gamers may be able to employ such emotive terms with the clinical detachment of the linguist, but to admit them to children is undesirable as putting a gun in their hands.

The other reason for prohibiting slang, colloquialisms and the like is that they tend to be too ephemeral for inclusion in the dictionary, which is why nearly all dictionary words designated "slang" seem so quaint and out of date. The reason is only valid if strict adherence to the dictionary is regarded as a rule of the game. I see no reason to prohibit words known to everybody merely because they are too ephemeral to be listed. This problem is overcome by my Rule of Appeal to other players.

uplink uplink

Technical terms

I (still) wonder how long it will be before byte ceases to be technical and acquires metaphorical value in everyday use. There is no reason to ban technical terms, even if not listed by the dictionary, provided that their existence can be proved or at least one other player acknowledges a claim. Their situation is, in fact, similar to that of slang words and colloquialisms. Note that words designated "foreign" in the dictionary may often be used in a technical sense - indeed, it is unlikely that they would otherwise be listed.

uplink uplink

Abbreviations and acronyms

I have met word-gamers who object to bus and phone on the ground that they are abbreviations and should be preceded by an apostrophe, and I once exchanged argument with a publisher whose copy-editor disinfected my use of ad by soaking it in inverted commas. In principle, no objection can be made to abbreviations in everyday use. In pract, howev, the princ is ope to abuse. You have no choice but either to put it to the vote or to see how it is listed and described by the dictionary.

It may be helpful to draw a distinction between oral contractions and written abbreviations. No objection can reasonably be raised against such natural spoken contractions as bus, phone, ad and suchlike. That they are derived from longer words is just a matter of history. Many words in use today have similar backgrounds, which their users are quite unaware of. It would be as pedantic to object to bus from omnibus as it would to reject mob on the ground that it derives by contraction from the two-word Latin phrase mobile vulgus. Written abbreviations, however, are merely orthographic conventions for words which are still normally pronounced in full, such as Mr for mister, Ms for manuscript and St for either saint or street. Abbreviations of this sort are clearly not allowable.

There are, inevitably, awkward borderline cases where pronunciation may follow orthographic abbreviation. For instance, the British National Scrabble Championship law-givers have accepted the word viz, a printer's conventional abbreviation of the Latin videlicet. Since what started as a purely orthographic convention has become "live" in many speakers' mouths, I welcome the acceptance. Unfortunately, none of the other members of my word-gaming circle have ever allowed me to count it, although I use it in speech, and I now see it is not listed in OSW.

We live in an age of acronyms. The anti-smoking lobby in Britain calls itself Action on Smoking and Health in order to secure for itself the appropriate acronym ASH, which strikes me as rather more successful than SIGMA for the Society of Inventors of Games and Mathematical Attractions. Examples of words that originated as acronyms include radar (for RAdio Detection And Ranging - note the additional appropriateness of the palindrome) and laser (for Light Amplification by Stimulated Emission of Radiation). Will LEM (Lunar Excursion Module) follow in their wake? Try using it in a word game and see. I would not object to it, as it bears all the hallmarks of a pronounceable English word - which is more than can be said for VTOL (Vertical Take-Off and Landing). Had they called it VERTAKOL, or even VERKOL, word-gamers would probably have found it acceptable.

In deciding whether or not to accept a claimed acronym as a word, it seems reasonable to apply two tests. First, is it generic or specific? UNESCO is essentially a name, therefore unacceptable if names are prohibited (as I believe they should be), whereas LEM could refer to any old lunar lander and for that reason is surely generic enough for acceptance.

Second, does it make an unambiguously pronounceable word according to the (admittedly somewhat haphazard) rules of English orthography? LEM is also acceptable by this test, but VTOL is not: VT never occurs at the front of an English word, and readers would vary in their pronunciation according to the vowel sound they insert between the two consonants. I would accept an acronym that passes both tests - which is why I never cease to be surprised that awol (Absent WithOut Leave) is not listed in either Chambers or the OED - it's a word that I use in speech and would accept as valid without hesitation.

uplink uplink

Hyphenated and compound words

A compound word is made when two or more are stuck together to make a whole that means more than the sum of its parts: a workshop is a place where a craft is carried out and would not do for the local shop where local shop-assistants work. The trouble with English is that we hyphenate some such compounds and not others. No player would object to workshop as a word claimed in a game, but I doubt that any would accept shopassistant. The artificiality involved in sticking words together or keeping them separate may be demonstrated by a phrase such as "State Security Head Office", which is a word-for-word translation of the c. 1939 German Reichssicherheitshauptamt. Word games tend to be tests of the written language, and what is or is not acceptable has nothing to do with sense or logic, only with form and convention. Whereas the Germans write their compounds as one word, we tend to keep them apart unless this leads to misunderstanding.

British word-players also suffer from hyphens, a disease from which Americans are comparatively free. Even where compounds are regarded as greater than the sum of their parts, the English generally prefer to keep them at a hyphen's length apart. To add to the confusion, the hyphen is also a piece of punctuation rather than an integral part of the spelling, and may be inserted or omitted according to the sense - a well-known fact that may not be quite as well known as it appears.

As Sir Ernest Gowers wrote in The Complete Plain Words, "If you take hyphens seriously you will surely go mad."

In British English we stress the noun rather than the adjective preceding it, as in red CABbage and red ADmiral. When such word pairs become more closely linked in a perhaps metaphorical sense, we may go so far as to hyphenate them in writing without altering the stressed syllable in speech, as in red-BLOODed or red-HANDed. Eventually, however, we so far regard the pair as forming a unit as to transfer the stress to the adjective before it as if the whole thing were a single noun, as in RED-belly and RED-head. A final stage of coalescence is reached when the hyphen is dropped from the written word, as in REDskin and REDcurrant. (Similar remarks apply to other word pairs such as noun-noun, object-verb and so on.) Looked at in this way, the presence of a hyphen in a compound may suggest that the words involved are undergoing a process of evolution into one.

Here are some further examples (hyphenation etc. follows Chambers in all cases):

Stage 1Stage 2 Stage 3Stage 4
long oddslong-windedlong-handlongbow
under ageunder-privilegedunder-trickunderdog
house ruleshouse-surgeonhouse-leekhousehold
iron lungiron-greyiron-mineironstone
round offround-tableround-uproundhead

In British word play, stage 1 compounds would certainly not be accepted, those at stage 4 certainly would be. Most would also disallow all those written with hyphens, although, on appeal from a claimant, I would allow those which could be regarded as belonging to stage 3.

This is one way of attacking the problem, but, of course, it is much too awkward for average word play. Prohibiting hyphenated words without examination is a practical necessity rather than a desirable ideal.

Americans do not have this problem, since the tendency of the language is to throw the stress back to the adjective or attributive regardless. "He's a crazy man", pronounced by Americans, usually sounds as if crazyman is all one word. Hence, in American, word-pairs may bed down overnight without passing through the formality of a hyphen. Words which are still single in British but indissolubly spliced in American include alright ("all right" in British), underway, anytime, anymore, worldwide and nodoubtmanyothers.

Some words, though hyphenated for one reason or another, are so patently single words that, if it were up to me, I would allow them in word games. Examples include X-ray, so-called, cul-de-sac, forget-me-not, and, especially, co-operative or co-operation. "If you must split them", says Gowers, "use a diæresis - that's what they're for." Unco-operative looks like a Scots expression; uncoöperative is better, and uncooperative better still.)

Whether or not to allow words containing apostrophes, such as fo'c'sle, is a matter of taste. But no one, I think, would argue in favour of two-word contractions like shan't, ain't or would've.

uplink uplink


Ah! and Eh? are interjections recognized by Chambers and most word-gamers. Chambers also recognizes Sh! and St!, which some players would feel dubious about if only because they don't contain vowels, along with Fy! and La!, which cannot have been heard in real English speech for a long time now. It does not recognize Aw! for a characteristically American form of self-expression, or Mm! for delighted agreement, or the equally conventional Zz! indicating mid-snooze, or even the less arguable and self-explanatory Ow!

If your dictionary is as inconsistent as Chambers in its choice of interjections, you have several choices of action. One is to accept all interjections listed plus those accepted by at least one other opponent on appeal. Another is to ban all words existing only as interjections, for which you would have my support. Yet a third would be to ban all two-letter words from word games and so dispense with almost the entire problem. That is my preferred solution, though I realise it would ruin board games like Scrabble (and Upwords, which I much prefer.)

uplink uplink

Two-letter words

There follows below a list of two-letter words that may be claimed but not necessarily accepted in word games. It is not intended to be helpful, but rather to support my preference for banning them completely, or at least restricting them to the 24 admitted into Ogden's Basic English, namely am an as at be by do go he ho if in is it me no of on or so to up us we. (Ogden doesn't actually include an, but it would be silly to omit it. Here's a link to his original text.)

As to the rest, dictionaries are generally inconsistent about what to include or omit, and whether or not a claimed word will be supported by the dictionary referred to is either a gamble or a question of list-memorisation. I recommend accepting or rejecting whole classes of words rather than individuals. If one sol-fa note is allowed, all should be; if one letter of the Greek alphabet, then all of them - together with those of the Roman alphabet, of course - and, while we're about it, why not also Hebrew?

The Oxford English Dictionary lists (by my count) 134 two-letter words, namely -

aa ab ad ae af ag ah ai al am an ar as at aw ax ay ba be bi bo by ce ch co da de do dy ea ed ee ef eh el em en er es et ex fa fe fy go gu ha he hi hm ho id if in io is it jo ki ko ky la le li lo ly ma me mi mm mo mu my na ne no nu ob od oe of oh oi om on oo op or os ou ow ox oy pa pe pi po qi ra re ri sa se sh si so st su ta te ti to ty ug uh um un up ur us ut uz ve wa we wo xi xu ya ye yi yo yu zo

Of these, the British Collins' Scrabble Wordchecker lists 113, omitting af ce co dy fe ki ly ra sa se su ty uz wa yi and acknowledging de as "US only", while Hasbro's American equivalent lists 96, omitting af ce co da dy ea ee fe fy gu io ki ko ky ly ob oi oo po ra sa se st su te ty ug ur uz wa yi yu zo.

Wikipedia no longer carries a list of two-letter words, but it did offer others such as the online encyclopedia, offers also av a month of the Hebrew calendar, di plural of Latin deus (true, but if you're going to allow Latin two-letter words why omit all the others?), gi a judo or karate costume, xe alternative spelling of sie, zi a type of pseudonym used in Imperial China. It also lists a number of other name words spelt only with a capital initial. (And, more amusingly, a second definition of ri as "a mythical creature said to be half human and half jellyfish".) There's a more up-to-date list at

Many of these words may be objected to on the following grounds:

Interjections. ag is South African, fy archaic, nu rare, ug dialectal, ur rarely so spelt (more usually rendered er, as is only human) and wo obscure, representing a call to a hawk or falcon, or, to a horse, a variant of whoa. Personally, I would allow ee, representing a Lancashire utterance (George Formby films still crop up on TV), which OED does not record as an interjection (but does as dialectal for eye), and zz as a conventional expression of sleep, for which, however, OED gives only ziz(z).

Foreign. de ki ko li ly ne oe om ti xu yu zo. American rules apparently permit use of French words de and ne (properly ), neither of which would I accept, and of course the spelling ax for what normally appears in British as axe. When not meaning a Faeroese whirlwind, oe is an English dialectal word for grandchild.

Dialectal. OED so designates ae af ch co ea ee et ky na oe oy su ug un uz ve wa zo. Of these, uz and zo are "Zummerzet" representations of West Country speech, and many of the others are not only dialectal but also obsolete.

Letters of the alphabet. ar ce ef el em en es ex mu nu pe pi xi. Four of these are well-known Greek letters used in mathematics. American sources accredit pe as a Hebrew letter, but OED recognizes it only as an obsolete variant of pee meaning a coat of coarse cloth.

Abbreviations. If you object to abbreviations, out go ad bi ed ex mo.

Archaisms and rarities. ab al ed ne ob.

Tonic sol-fa. Everybody knows the usual ones - ut re mi fa la te - and perhaps also si as a variant of the leading-note te. Less well known is the fact that sharps are expressed by changing the vowel to "e", so producing an additional series de re me fe se le, and flats by "a" (pronounced "aw"), extending it yet further to da ra ma fa sa la ta.

Copyright 2004-