A Happy Ending

A few months ago we received a communication from astute Lounge visitor Mr. Wood Foster-Smith, who wrote to us as follows:

...words ending in -id and -or: I've noticed that there is a wonderfully descriptive array of adjectival words ending with the -id suffix (vapid, turgid, putrid, gelid, tepid), many of which have noun counterparts ending in -or (stupid/stupor, squalid/squalor, fetid/fetor, candid/candor, torpid/torpor); but there are also false pairings of these words (tumid/tumor, rancid/rancor, valid/valor) and -or words without an -id counterpart (ardor but not ardid, honor but not honid), as well as the other way around (acrid but not acror, florid but not floror). How did these words come to be? Are they related in some morphological sense? Are the false pairings above actually etymologically related, but their meanings have diverged over time? I have always loved these words and would love to hear some erudite ruminations on them.

First of all, "Well said!" we cry. And while we cannot promise erudition, we concur that these word pairs constitute an interesting pattern for which there is, of course, an explanation. We relish the opportunity to provide it, as it justifies one of our favorite harmless activities: wandering through the dusty old grammar books that fill the walls of the Lounge Library.

To begin: it is always worthwhile keeping in mind a hard truth about language generally, and particularly about English. It is well expressed by the great usage authority H.W. Fowler, even though he was commenting on a rather small area of English when he wrote it. He said that some relations among words

come to us from our forefathers as an odd jumble and plainly show that the language has not been neatly constructed by a master builder who could create each part to do the exact work required of it, neither overlapped or overlapping; far from that, its parts have had to grow as they could.

And so it is with the group of adjectives and nouns in question, as well as others we'll look at.

English adjectives ending in -id that have related nouns ending in -or (or, as the Brits have rather quaintly maintained through the centuries, nouns ending in -our) are all Latin derivatives. The Latin adjectives end in -idus, the nouns in -or. For the most part, these pairs also have a common verb at their root, thus explaining the existence of the many pairings that Mr. Foster-Smith refers to. To quote a few august authorities (in which we have highlighted some points in color, for reasons soon to be explained):

From A Latin Grammar for the Use of Schools (1888):

-idus ...is affixed chiefly to the stem of intransitive verbs in -eo, [and] denotes the condition and property which are expressed by the verb... Some few are formed from other verbs or from substantives, or have no known primitive.

From Allen and Greenough's New Latin Grammar for Schools and Colleges (1903):

Adjectives expressing the action of the verb as a quality or tendency are formed from real or apparent verb-stems with the suffix [-idus. It] is no doubt denominative, as in herbidus, grassy (from herba, herb); tumidus, swollen (cf. tumu-lus, hill; tumul-tus, uproar); mucidus, slimy (cf. mucus, slime)... But later it was used to form adjectives directly from verb-stems.

From "Latin -idus and -tudo" in the journal Language (1946):

The Latin adjectives ending in -idus generally appear beside intransitive verbs of the second conjugation and abstract nouns ending in -or, -oris: e.g. timeo, timor, timidus. They represent, as it were, the participle... The meaning of the adjectives does not require a derivation from the verb associated with it: in many cases the derivation of the verb from the adjective is clear.

With apologies to readers who are now gasping and shrieking "TMI! TMI!", we note the upshot of our color highlighting: even in Latin the relations among these adjective/noun pairs and their roots were not entirely consistent, so it is only natural that we should have inherited some defective pairs in English. This explains, to some degree, the observation that English has some -id adjectives without a corresponding -or noun, and vice versa.

As for the "false pairings" that Mr. Foster-Smith noted: these particular pairs are in fact not false; they are etymologically related but have diverged enough in meaning that their relationships are perhaps not immediately obvious. Tumid and tumor share an ancestor (and meaning) in the sense of "swell"; they have another Latinate English relative in the adjective tumescent. Rancid and rancor share an ancestor in a Latin verb meaning "be rotten," but rancor in English today nearly always applies to a psychological, rather than a sensate quality¹. Finally, valid and valor have a common ancestor in the Latin verb meaning "be strong, be of worth." They have an English cousin in value.

A second source of possible confusion in -id adjectives is the fact that English generally is just wild about the word ending -id. It makes other appearances in English words that have no connection with Latinate -or words. Here are some of the patterns, starting with our very favorite:

1. Adjectives and nouns ending in -id related to adjectives and nouns ending in -ine, all of which are what the OED refers to, charmingly, as "zoological appellatives." Case in point: bovid (a member of the cattle family), and bovine (relating to cattle). This set of pairs (many of which hold hands in the VT) is quite large and nearly nondiscrepant since all the words are derivatives of fully formed Latin inflections. There are, however, many oddly imbalanced such pairs in which one form is used far more than its counterpart: canine and feline, for example, far outnumber the rarely seen canid and felid.
2. Somewhat related to these nouns and adjectives are words denoting characteristics of a line of descent (such as hominid, mongolid). These adjectives occasionally have noun counterparts ending in -ian, many of which are obscurely technical or obsolete (like hominian).
3. Latin nouns (many derived from Greek inflected forms) with the terminal spelling -ides which English writers sensibly truncated to -id. Again, a large list (orchid, pyramid, chrysalid, carotid), that does not have a consistent relationship with another English word form.

There are, then, many groups of -id words for you to sink your bicuspids into (including the names of mammalian teeth). What's the take-home lesson? It's a lesson that's been kicking around for centuries, and that refuses to die, despite all the bells and whistles of modern education: there's nothing like the study of classical languages to clue you up about English word formation.

Here's a page where you can bone up on your -ine animal adjectives and acquire many new ones that will amaze your friends:

http://www.phrases.org.uk/animal-adjectives.html

Fowler's Modern English Usage (quoted above) is, to our minds, one of the most delightful books ever written in English. With due respect to the editor of the current 3rd edition, we prefer the 2nd edition, still widely available in book form and also online:

http://www.bartleby.com/116/

You can get a taste for the great master's erudition, wit, and uncompromising good sense by sampling the beginning of Chapter III, "Airs and Graces":

http://www.bartleby.com/116/302.html

¹Similarity to the adjective rank (meaning "offensive in smell or taste") is a red herring; that word is from Old English.