When I use a word . . . The longest medical words

A linguistic paradox

Some words and phrases describe themselves, while others do not.

“Short,” for instance, is a short word, but “long” is not a long word.

“Polysyllabic” is polysyllabic, but “monosyllabic” is not monosyllabic.

“Pronounceable” is pronounceable, but “unpronounceable” is not unpronounceable.

“An English phrase” is an English phrase, but “a French phrase” is not a French phrase.

Let’s give names to these two different types. Let’s use the word “homological” to describe words and phrases that describe themselves and “heterological” to describe those that don’t describe themselves.

Now we can ask whether these two words do or do not describe themselves.

“Autological describes words and phrases that describe themselves. So “autological” is autological.

However, if “heterological” describes itself then it is not heterological, and if it doesn’t describe itself then it is heterological. In other words, as it has been expressed elsewhere, in more formal logic terms1:

“is heterological” is heterological

is true if and only if

“is heterological” is heterological.

is false.

In other words, “heterological” is heterological only if it isn’t.

This paradox is called Grelling’s paradox or the Nelson–Grelling paradox or antinomy.2 And, as the title of the original paper by Grelling and Nelson implies, it was inspired by considering paradoxes named after Bertrand Russell and Cesare Burali-Forti.

Russell’s paradox involves the idea of classes or sets of objects. Most sets are not members of themselves. The set of all languages, for example, is not itself a language. But a set of sets, being itself a set, belongs to itself. However, we can also make a set of all non-self-membered sets and then ask whether that set is a member of itself. And then we realise that if it is a member of itself it’s non-self-membered and therefore isn’t a member of itself; and vice versa.3

Burali-Forti’s paradox, which stimulated Russell to formulate his paradox, is a similar paradox involving sets of all ordinal numbers.4

These three paradoxes all arise from self-reference. Other examples of self-referential paradoxes include the liar paradox, which arises from the statement, by a Cretan, that all Cretans are liars, and Berry’s paradox, which arises from a statement describing a particular integer, namely “the least integer not describable in fewer than twenty syllables,” which is a self-referential statement that is only 19 syllables long.5

What makes a word long?

If we agree that “short” is a short word, we can hardly claim that “long” is a long word. So what makes a word long? “Longer” is longer than “long” and “longest” is longer still. So “longest” is the longest word in the triplet, “long,” “longer,” and “longest.” But it’s not as long as, for example, “longhand” or “longevity” or “longaevous” or “longanimous” or “longitudinal” or “longinquities” or “long-windedness” or “long-sightedness.” And that’s the long and short of it.

Here another paradox may help, the sorites paradox. The word “sorites” comes from Latin via the Greek word σωρός, a heap, and the difficulty the paradox poses arises from heaps.

Imagine a heap of 1124 tennis balls arranged in a pyramid of 16 layers, with 192 on the bottom layer, 169 on the second layer, 147 on the third, and so on, up to three on the second top layer and one at the very top. Now take one away from the top, leaving 1123. It’s still a heap, a truncated pyramid. Now take more balls away, one at a time. How long does the heap remain a heap? You would be able to make a mini-pyramid from the last four balls, with three on the bottom and one on top, but would you still regard it as a heap? What about 11 balls arranged in a pyramid of three layers?

Now imagine doing all this with grains of sand instead. How many would you be able to make into a recognisable heap as the number gets smaller? Where is the exact cut-off point beyond which the heap disappears?

Now consider a different problem, one that still involves subtraction a little at a time. You have a glass of water at 40°F. You lower the temperature one degree at a time, and when you reach 32°F the water turns to ice.

In both cases you removed one unit of measurement at a time, but in the second case the physical change that occurred allowed you to recognise a definitive cut-off point that had no discernible analogueue in the first case.

Just as there is no clear cut-off point that determines when a heap of grains of sand is no longer a heap, so there is no cut-off between short and long words. However, in its absence an artificially created one may help. This leads me to propose two criteria for distinguishing long words from short ones.

● Criterion 1. Long words should have at least 10 letters in them. This is an arbitrary cut-off point; others might choose a different one.

● Criterion 2. Nevertheless, even if you accept my first criterion, it is insufficient. Consider the nonsense song “Supercalifragilisticexpialidocious,” in which the two longest words, apart from the title itself, are “atrocious” and “precocious.” The first has nine letters and the second 10, but the second doesn’t really feel longer than the first, contradicting Criterion 1. Something more is needed, and that something more is the number of syllables, each having only three. My second criterion, therefore, is that a long word should have at least four syllables. Again, an arbitrary cut-off point.

When is a word not a word?

As soon as one tries to find very long medical words, one runs into difficulty. The problem is that all the longest words you can find don’t really qualify as words at all.

The word “word” has too many definitions listed in the entry in the Oxford English Dictionary (OED) to recount here,6 and in any case doing so wouldn’t help me explain what I want to explain. You have to work your way down to section III.12 of the entry before you get to the helpful part. Section III is headed “An element or unit of speech, language, etc.” and section III.12, which surprisingly is the first part in section III, tells us that a word is “Any of the sequences of one or more sounds or morphemes (intuitively recognized by native speakers as) constituting the basic units of meaningful speech used in forming a sentence or utterance in a language (and in most writing systems normally separated by spaces); a lexical unit other than a phrase or affix; an item of vocabulary, a vocable.”

The dictionary then adds: “Sometimes used specifically to denote either an item of vocabulary in the standard form in which it is generally cited in a dictionary, etc. (e.g. the infinitive of a verb), or this form considered together with its grammatical inflections as expressing a common lexical meaning or range of meanings.”

From all this, being “a native speaker,” I take it that a word is a pronounceable concatenation of letters representing a concept, such as an item or an action, that conveys meaning to me, and furthermore that the meaningful concatenation should be discoverable in a standard dictionary. I say a “standard dictionary” because there are dictionaries that are by no means standard, which sometimes include items that I would not regard as words under this definition.

Here’s an example. Many online sources refer to a “word,” 18 819 letters wrong, that describes the chemical structure of a protein, titin or connectin, present in skeletal muscle. Its name starts “methionylthreonylthreonylglutaminylalanyl....” and ends “...isoleucine.” It’s said to be the largest known protein, and it may well be, at least until a larger one is discovered, but the string of letters that represents the string of amino acids from which the protein is formed is the name of a chemical formula, which can hardly be regarded as a word.

There’s another similar example in Mrs Byrne’s Dictionary of Unusual, Obscure and Preposterous Words (no, I’m not making it up). Josefa H Byrne was the daughter of the great violinist Jascha Heifetz, and she spent some time combing through various sources finding material for her dictionary, which can’t be regarded as “standard.” In her collection she mentions a polypeptide enzyme whose name contains 1913 letters, consisting of the string of terms used to describe each of its 267 constituent amino acids; it starts “methionylglutaminylarginyl.....” and ends “....threonylarginylserine.” This one is better known as tryptophan synthetase. But is it a word? Not one that I’ve used recently.

Leaving aside chemical monstrosities of this sort, we find a variety of considerably shorter, but still lengthy, concatenations specifically devised for humorous or other literary purposes. For example, in his novel Headlong Hall (1816) Thomas Love Peacock created a 51-letter chain to describe the different constituents of the human body. It comes in Chapter XI (The Anniversary) and the words are attributed to Mr Cranium: “Ardently desirous, to the extent of my feeble capacity, of disseminating as much as possible, the inexhaustible treasures to which this golden key admits the humblest votary of philosophical truth, I invite you, when you have sufficiently restored, replenished, refreshed, and exhilarated that osteosarchaematosplanchnochondroneuromuelous, or to employ a more intelligible term, osseocarnisanguineoviscericartilaginonervomedullary, compages, or shell, the body, which at once envelopes and developes that mysterious and inestimable kernel, the desiderative, determinative, ratiocinative, imaginative, inquisitive, appetitive, comparative, reminiscent, congeries of ideas and notions, simple and compound, comprised in the comprehensive denomination of mind, to take a peep with me into the mechanical arcana of the anatomico-metaphysical universe.” I couldn’t have put it better myself. It’s impressive and a pretty comprehensive medical description, but it’s hardly a proper word, and you certainly won’t find it in any dictionary, not even Mrs Byrne’s.

On the other hand Mrs Byrne does include a 182 letter transliteration of the 170 letter Greek word used by Aristophanes in his play The Ecclesiazusae to describe the leftovers of last week’s meals. Here’s the Greek version:

Λοπαδο-τεμαχο-σελαχο-γαλεο-κρανιο-λειψανο-δριμ-υποτριμματο-σιλϕιο-καραβο-μελιτο-κατακεχυμενο-κιχλ-επικοσσυϕο-ϕαττο-περιστερ-αλεκτρυον-οπτο-κεϕαλλιο-κιγκλο-πελειο-λαγοιο-σιραιο-βαφη-τραγανο-πτερυγον;

...and here is the English transliteration:

lopado-temacho-selacho-galeo-kranio-leipsano-drim-hypotrimmato-silphio-karabo-melito-katakechumeno-kichl-epikossupho-phatto-perister-alektruon-opto-kephallio-kinklo-peleio—lagoio-siraio-baphē-tragano-pterygon;

...and here’s what it means:

a platter of sliced shark and dogfish skull, with a leftover pungent sauce of silphium, plus crayfish, honey-basted thrush, blackbird, ringdove, pigeon, and cock, with roasted mullet, wagtail, rockdove, and hare, plus red wine and mulled giblets and wings.

I wonder if Kenneth Grahame was thinking of Aristophanes when he described the contents of the picnic hamper, 108 delicious letters, that Rat and Mole took with them on their river expedition in the first chapter of The Wind in the Willows (18xx): Coldtonguecoldhamcoldbeefpickledgherkinssaladfrenchrollscresssandwidgespottedmeatgingerbeerlemonadesodawater. After all, he may well have been familiar with Liddell and Scott’s Greek–English Lexicon (1843), in which, typically thorough, they included Aristophanes’ monstrosity, although without transliteration. Incidentally, I’ve had to correct an error in Mrs Byrne’s transliteration—she doesn’t know that when the Greek letters gamma and kappa come together, as γκ in κιγκλο, they are idiosyncratically pronounced as “nk,” not “gk,” i.e. kinklo, not kigklo.

Now you may be thinking that Liddell and Scott, as it is usually known, is a standard dictionary, and when it comes to bilingual dictionaries, it is. But not the kind of standard dictionary I have in mind, i.e. a monoglot English dictionary, such as the OED.

And you might also be thinking that for proper words, as I should prefer to define them, you should look in such standard dictionaries and find the longest. Unfortunately, it’s still not so straightforward.

The longest word included in the second edition of the OED is a medical one, 45 letters long: pneumonoultramicroscopicsilicovolcanoconiosis. And it’s easy to work out what it means. When Greek athletes contested in the nude they oiled their bodies. Wrestlers then applied κόνις (konis), a fine dust that gave them a better grip on each other. So a coniosis, like pneumoconiosis, is a disease caused by dust. In this case the dust is deposited in the lungs πνεύμων, pneumōn), is very fine (ultramicroscopic), and is derived from volcanic silica.

Alas, this delightful word turns out not to be a proper word at all, being defined in the OED as “A word invented (probably by Everett M. Smith (born 1894), president of the National Puzzlers' League in 1935, in imitation of polysyllabic medical terms, a factitious word alleged to mean ‘a lung disease caused by the inhalation of very fine sand and ash dust’ but occurring only as an instance of a very long word.”7 Since then a supposed case of the disease has actually been described,8 an unusual example of the invention of the name of a disease before its clinical description. So perhaps it’s become a word after all.

A final thought

Having defined a long word I have tried to find some medical examples, but my efforts have been thwarted by the occurrence of concatenations that cannot be properly described as words at all, in the usual sense of the word.

So, in order to find some very long medical words, I shall have to try harder and perhaps set my sights lower regarding the lengths of the words I seek.