Writing about ghost words got me thinking about collective terms for animals, especially the weird ones that someone always trots out as a bit of trivia. (I once won a barroom trivia contest for being able to state that a group of rhinoceroses was called a "crash.")

The thing is, apart from such occasions, when does anyone actually use those words? And, more importantly, how often has anyone ever used those words? The questions are relevant, because lists of collective terms for animals can be found all over the place, and as decontextualized lists, they take on the mantle of an authoritative pronouncement: this is the term for X; if you want to appear erudite, use it.

But what support do these words actually have? Many of the words in these lists look like they come from a parlor game: think of a term for an animal that expresses something about a putative characteristic of the animal in question. So if we associate crows with death, voila, "a murder of crows." In fact, many of these collective terms for animals were popularized by James Lipton in his book An Exaltation of Larks, in which he explains this "game of venery." Since all the lists floating around seem to derive, directly or indirectly, from Lipton, who explicitly plays this game, we need to be cautious. Frankly, people who express a fondness for obscure words are often just too credulous, too willing to accept an assertion about a word's meaning just because it seems exotic, striking, or old.

Let me ask you to do a thought experiment. Suppose that I play the game of venery and invent a collective term for logophiles (word-lovers): a credulity of logophiles. Are you going to accept this as the "proper" collective term for this group? I doubt it. Now suppose that someone played this game 500 years ago and wrote down similar terms. Suppose further that this list was buried in a library for hundreds of years until some industrious person discovered it, edited it and printed it. Does the patina of age suddenly make these words "proper", even though no one has used them since the time they were created?

The basic problem here is deciding what makes a word part of English. True, one cannot set precise boundaries (we have a fuzzy category at issue), but surely nonce-formations do not have a claim to anything more than a shadow existence on the margins of the language, at least until enough speakers or writers adopt the term. Before that point, a coinage may be witty, humorous, and intriguing, but it has no normative force. In other words, you can't expect anyone to accept a one-off use as standard English.

And what about different dialects. It's quite different to say that X is a group term in a specific regional dialect and to claim status for the word in general English. These lists make no such distinctions.

Now many of the collective terms in Lipton et alia are indeed old. More than a few date back to the 15th-century Book of St. Albans. But look them up in the OED. Over and over, you find definitions expressing skepticism about the word's validity: "supposed term for...", along with a citation either only to that first word list, or to a series of word lists that all copied each other.

To find out how many of such words have any authority for their existence prior to Lipton's book, I looked up the putative collective terms for the list of animals found in The Insomniac's Dictionary: The Last Word on the Odd Word, by Paul Hellweg. This is a shorter list than Lipton's, but still, I think, representative of the sort of thing one finds.

It's quickly apparent that there are problems with many words in the list. Nineteen of them don't appear in the OED in any collective sense at all.

bale of turtles
chattering of starlings
clamor of rooks
cowardice of curs
crash of rhinoceroses
deceit of lapwings
descent of woodpeckers
murder of crows
mustering of storks
ostentation of peacocks
paddling of ducks
piteousness of doves
pitying of turtledoves
rafter of turkeys
richness of martens
route of wolves
smack of jellyfish
spring of teal
tidings of magpies

Five more not only don't have a collective definition that fits, but they appear to be errors based on misreading attested definitions:

dray of squirrels
harras of horses
hover of trout
singular of boars
peep of chickens

A dray is a dialectical term for a squirrel's nest. Harras is doubly problematic. First, this form is an obsolete spelling variant. The OED's entry is under the form haras. Second, the term fundamentally means "an enclosure or establishment in which horses and mares are kept for breeding." As a result of that basic meaning, it is also used (obsoletely) in the extended sense of "a stud, breed, or race of horses." Notice, please, that even this obsolete meaning is not a collective. In other words, it's not equivalent to a herd horses. It was used to talk about their lineage (e.g., this horse comes from a fine haras).

The only connection listed between hover and fish is as a term for a sheltering overhang under which fish can hide. There is one attestation of "a hover of crows", i.e., a hovering host. I suppose this is a better candidate for a collective than "murder," but it can well be explained as a poetic nonce-formation rather than a regular term. Singular does appear in the venery lists, but it's a variant of sanglier, a name for a grown boar. Did no one ever pause to wonder about the illogic of calling a group of something a singular and stop to check it out! In a similar mistake, a peep is a name for an individual young chicken (or other bird).

The OED is not, of course infallible, and it's possible that the compilers missed a few, but as entries below show, they did consult The Book of St. Albans and other word lists. It seems highly implausible that they missed all these. Even if a few were oversights, all these terms can collectively have had essentially no impact on the language if the most comprehensive dictionary in existence missed out on them.

Sixteen more words do appear in the OED, but the only citations are to the word lists. In other words, they appear to have no independent existence in this sense. Many of these are defined with qualifications like "A pretended term for..." that express the lexicographers' skepticism.

barren of mules
building of rooks
business of ferrets
clowder of cats
fall of woodcocks
host of sparrows
labor of moles
leap of leopards
muster of peacocks
pace of asses
parliament of owls
rag of colts
shrewdness of apes
sleuth/sloth of bears
unkindness of ravens
walk of snipes
watch of nightingales

Although clowder in this form only appears in a word list, it is a variation on cludder, a dialectical term for "cluster." In that form, it has a number of citations for groups of things (but none for animals). The expression "parliament of rooks" is also cited, although the context seems figurative, cf. Chaucer's poem Parliament of Fowls rather than a generic collective. The only citation for "parliament of owls" is to Lipton. Also note that I'm counting sleuth and sloth as variant forms of the same term.

One item has a prior use outside the lists, but only in the sense of the young of any animal, and is not used collectively except in the word lists:

kindle of kittens

Four terms originated with the word-lists but managed to make the leap into somewhat wider currency.

exaltation of larks
gaggle of geese
murmuration of starlings
skulk of foxes

Of these four, only gaggle has a real claim to be standard. Although it's not in Hellwig's list, pride is another term which originated in the word lists but which has caught on as a general term, although it's interesting that apart from the Book of St. Albans all the citations are 20th century. Exaltation, which has very little currency before Lipton, I am giving the benefit of the doubt by putting here. The earliest citation predates the word lists, but is from the Lydage poem The hors, the shepe, [and] the ghoos, which is probably a source for some of the Book of St. Albans, and which is definitely fanciful rather than reflecting accepted usage. The handful of later citations don't come from word lists, but they appear to be a reaction to the word lists. For example:

1824 J. McCulloch Scotland III. 407, I have never spoken of 'an exaltation of larks'.

Murmuration was dormant until the 20th century, when it was modestly revived. The most notable citation is to Mervyn Peake's Titus Groan.

Finally, we come to those words that have some defensible use as collectives. Three of them can be used as general terms for groups, but there is no citation for the animal collective:

convocation of eagles
knot of toads
party of jays

For party, there is a citation "I have found these birds in little parties in the autumn..." but that is clearly just a personified use of the ordinary meaning.

Ten terms have a legitimate existence as collectives, but need qualification, because their actual definitions are slightly different than is implied by a simple list:

bouquet of pheasants
cast of hawks
colony of penguins
company of parrots
flight of swallows
gang of elk
skein of geese
trip of goats
troop of kangaroos
wedge of swans

A bouquet only applies to pheasants in a specific context: "The flight of a multitude of pheasants breaking covert from the central point at which the beaters meet." It can also refer to the central point itself. A cast refers to the number or of hawks or any other birds of prey cast off from the arm of someone hunting with birds. It frequently refers to a pair. Colony, company, and troop don't have specific citations for the animals named in the list but are terms whose basic meaning implies a group, and are much more widely used (e.g., a colony of bees, a company of fowls, a troop of monkeys, etc.). Because of that basic meaning, there's no reason to object to their use for these animals, but at the same time there's no reason to think that these terms have special status for these animals. A flight doesn't refer specifically to swallows, but to any group of birds that take wing at the same time. A gang can also refer to buffalo and other herd animals, and is a U.S. term, rather than something general to all English-speaking countries. Skein and wedge refer to any group of wildfowl in flight. Trip, which is a variant of troop, can refer to a small flock of goats, sheep, hares, etc.

Five terms are both accurate and owe their existence to something other than the word lists:

charm of finches
siege of herons
pod of seals (or whales)
gam of whales
sounder of swine

Siege is in the word lists, but has a prior existence.

One thing that is notable when you look at the words as grouped above, there is a very strong correlation between the "bunk" words and the game of venery, especially the tendency to pick terms that are supposed to be characteristic of the animals in question. And with a handful of exceptions that made the jump from the word lists into general use, the collective terms that people actually use (and have used), do not refer to characteristics.

The apparent counter-examples above have different etymologies than you might guess from their current form. For example, charm doesn't derive from the ordinary charm but is a variant of chirm, which means the noise of many birds, etc., and by transference, the group that produced the noise. And sounder has nothing to do with sound, but is related to the Old English sunor, which also means "herd of swine." Siege comes from the meaning of that word as a seat, specifically, a bird's perch.

To summarize, most of the words in this list are the result of a game. Mingled with those fanciful words are legitimate terms and outright errors. All of them are being passed off as "correct" terms. Now I would be the last person to deny that Lipton's words and the game of venery are fun. But it's a big leap to go from a word game to the assertion that these words are all "proper" English. Those who do so truly belong to a credulity.

[Note: after compiling this list, I found this list of bird terms which makes a similar classification of the various terms, although limited to birds.]