180

A CLASSiFICATION OF WORDS

MAXEY BROOKE Sweeny, Texas

One of the pleasures of the logologist is investigating the relations between pairs of words. Words can be characterized in four ways: spel­ling (or appearance), sound (or pronunciation) , meaning (or definition) , and origin (or derivation). Two words can be either identical or differ­ent with respect to each of these properties; thus, there are 16 different ways in which word-pairs can be classified, ranging from identical in all four respects to different in all four respects. The 14 intermediate cases, those in which a pair of words ·are identical in some respects but differ in others, are the ones of most interest to the logologist.

This work was inspired by Dave Silverman, who, in the May 1974 Kickshaws, introduced a slightly simpler 8- way classification, ignoring origin but distinguishing spelling, sound and meaning. He suggested homograph, homophone and synonym to describe word-pairs identical in exactly one of these characteristics, and gave examples (but not defi ­nite names) of word-pairs identical in two characteristics.

For brevity, I introduce the following notation. Let Sp stand for spel­ling, So for sound, Me, for meaning and Or for origin; also, let paren­theses enclosing any of these abbreviations denote that the two words are unlike in that respect. For example, word- pair s identical in spelling and sound but different in meaning and origin are denoted by the symbol­ism SpSo( Me) (Or) .

Over the years a considerable number of words have been introduced to distinguish words according to these properties. Unfortunately, the definitions do not always specify the presence or absence of each proper­ty; for example, two words are synonyms if they have the same meaning, regardless of whether or not they have the same or different sounds, spellings or origins. Worse, dictionaries often give different definitions for the same word. I summarize below the definitions given several common and not- so- common words by the Oxford English Dictionary (OED) , The Merriam- Webster Unabridged (W3, W2), and the Funk & Wagnalls Unabridged ( FW) , as well as by a dictionary devoted to lin­guistic s, Mario Pei and Frank Gaynor I s Dictionary of Linguistic s (Crown Publishers, 1954) ( PG) .

conjugate OrMe OED- W3- FW doublet Or OED- W3; (O)? FW; Or( Me) PG heteronym Sp( So) (Me) OED- W2- FW- PG homograph Sp( Or) (Me) OED- FV1; Sp( Or), Sp( Me), Sp( So) W3;

homomor: homonym

homophon

paronym polyonym synonym

It is clear fre ought to be a' nym is some' of homonym 1

He re is ]

So( Sp) ( Me) ( Gaynor, t have appe 1973) and ust 1969) homophol the same

Sp ( So) ( Me) ( tar). Ag graph; h~

ing With Me( So} ( Sp) (

that the v regardlel I propose

Or( So) ( Sp) ( ] this prec the word

SoSp( Me) (Or and SQU) Dave Sih shaws, tl tide in t] a word a]

SoMe(Sp) (01 ( German and prop'

SoOr ( Sp) ( ME kanna, IT

sense of sense of a quasi-l homopho:

Ie relations llr ways: spel­or definition) , lcal or differ­'e 16 diffe rent ~dentical in .ntermediate : respects 5"ologist.

: May 1974 ion, ignoring suggested s identical (but not defi­

stand for spel­), let paren­;wo words are in spelling '( the symbol-

en introduced lnately, the : each proper­:ame meaning, It sounds, mt definitions n several lictionary the Funk & ted to lin­llistics

Sp( So) W3;

181

Sp( Or) (So) (Me) PG homomorph = homophone W2 - PG homonym So( Me) OED; SpSo( Me) , = homophone, = homograph

W3; So(Me), Sp(So) (Or), So(Sp) (Or) FW; SpSo( Me) (Or) PG

homophone So ( Me) ( 0 r) • So ( 0 r) 0 ED; So ( Me) , So ( 0 r) , So ( Sp) W3; = homonym FW; So( Sp) ( Or) (Me) PG

paronyzn or OED- FW; So( Sp) (Me). = conjugate FW; Or( Me) PG polyonym = synonym W2- OED synonym Me OED- W3- FW-PG

It is clear from this that homonym is a hopelessly confused term which ought to be avoided in any systematic classification. Although hetero­nyzn is somewhat clearer, it is so frequently regarded as the opposite of homonym that it. too. should be eschewed.

He re is my proposed nomenclature for classifying pairs of words:

So( Sp) (Me) (Or) example, FLOCKS and PHLOX. Following Pei and Gaynor. this should be called a homophone. Various generalizations have appeared in Word Ways. such as queer reed/queried (February 1973) and new dandy canary's skin/nude Andy can air his skin (Aug­ust 1969). On the other hand, the August 1969 issue suggests that homophones can refer to different groups of letters in words having the same sound (as uff-uph).

Sp( So) ( Me) (Or) example. TARRY (to wait) and T ARR Y (covered with tar). Again following Pei and Gaynor, this should be called a homo­graph; heteronym is a less-acceptable synonym. In his book Play­in, With Words, Joseph Shipley calls these doubletones.

Me( So ( Sp) ( Or) example, PORKE R and PIG. All references agree that the word synonym describes a word- pair identical in meaning, regardless of any other characteristic. Since this is too general, I propose the following modification: definitive synonym.

Or( So) (Sp) (Me) example. GUARD and WARD. No dictionary defines this precise combination of characteristic s; I suggest borrowing the word doublet.

SoSp( Me) (Or) example. SQUASH (a vegetable, from the Indian asqnash) and SQUASH (a game, from the French esquachier). Described by Dave Silverman as words with two origins in the May 1969 Kick­shaws, these can be christened double origin words. Drury' s ar­ticle in the August 1969 Word Ways calls these words homographs, a word already used above.

SoMe( Sp) (Or) example. COLE (Anglo-Saxon for cabbage) and KOHL (German for cabbage). I can find no term for this classification, and propose synonymic homophone.

SoOr( Sp) (Me) example, CANON and CANNON, both from the Greek kanna. meaning II reed"; canon gets its meaning from reed in the sense of a rod, while cannon gets its meaning from reed in the sense of a hollow tube. Although this classification has been termed a quasi- homograph in the August 1969 Word Ways. I believe that homophonic doublet is a clearer description.

182

SpMe( So) (Or) example, QUAY (pronounced 1l kee ll , from the French) and QUAY (pronounced 11 cay", from the Spanish). I can find no term for this clas sification, and propose synonymic homographs.

spar( So) (Me) example, PRESENT (verb) and PRESENT (noun). I find no term for this classification, and propose homographic doublets.

MeOr(So) (Sp) example, KENTUM (Latin) and SATEM (Sanskrit), both meaning II one hundred". I find no term for this classification, and propose synonymic doublets.

SoSpMe( Or) example, YANKEE (claimed by Thomas Anbury in his 1789 work, Travels, to corne from the Cherokee eankke, or II coward '1 )

and YANKEE (clai med by H. L. Mencken in his 1937 book) The Amer­ican Langua~e, to corne from the Dutch Janke, the diminutive of "jan"). T ese are words that have conflicting derivations; the classic example is the word OK, discussed by Ralph Beaman in the May 1975 Word Ways. I propose the name derivation variants.

SoSpOr( Me) example, CONDUCT (to lead) and CONDUCT (behavior) . Although this classification has been termed a pseudo-homograph in the August 1969 Word Ways, I believe that meaning variant is a clearer description.

SoMeOr( Sp) example, CHECK and CHEQUE. These are merely spelling variants.

SpMeOr( So) example, CA RlBBEAN (accented on the third syllable) and CA RI~BEAN (accented on the second syllable). These are merely pronunciation variants.

Thus endeth my essay. I am indebted to Dmitri Borgmann for various suggestions and additions.

A 700-~

D. L. paLOI' Newtonville, ]

God, as a Here Hero 1 s 1

Late Madam I Eliot' s. Ah, ta sted acetic. Adam, or far' ma1s are odde Desi Arnaz, r is a bone, idi< led opals do g Is debris pure Away at Syria fad. P d keep pond, a hip, a A snide, viler wet sleet is in moot. Tattle Never on a do I won tubs. E zookeeper, at us similar, 0 gigolo. Nevel amiable. Mel erupt. Ah, tc plasma. Yet is no beep at ( I am no slob, Lear's Italiar. in a bog; no c Tin made fast late, big Abe, nis, go fall in anal pat. Car. erase song. Melba I made peg. A tire h rat sleep as I away as we fr Bedside tales lap Ode1 put c die. No, basi