FINDINGS ON THE OUTCOME OF ANALYSIS ON THE OED

A need for a categorization similar to the system used on the source compilations (idiom dictionaries) arose from

observations on the OED accounts. Hence, a new investigation using the same method and grading system was conducted for the assessment of data in the OED.

A surprising finding very soon cropped up: on an average, the details provided in the OED were uncommonly frequently found to be equally incomplete and defective - occasionally downright unreliable - as the information presented in the idiom compilations or dictionaries. The etymological record, it is true, offered in the OED with its illustrative quotations is, beyond doubt, the best it can offer on its 22 000 pages. Nevertheless, the fact remains that this treasury of historical record was found to fail its student on the aspects determined essential for this

investigation even more often than the sources.

This is unarguably corroborated by the following

comparative data: the sources included 73 proper name idioms in Category 3 vs. the 55 items to be found in the OED in this

category. In Categories 1 and 0 the two parties (i.e. source compilations and the OED) were practically of equal strength (differing in three cases only, to the discredit of the OED).

The only field where the latter excelled over the sources was Category 2, with 22 items against the 37 in favour of the OED. Nevertheless, to be impartial towards both, one does well to bear in mind that the former figures represent the prime of several comprehensive accounts to be found in the sources while the record supplied by the OED goes always back to the one and same work. As a rule, these OED defects were due to the following shortcomings: superficial treatment of entries, lack of discussion on rivalling etymologies,

fragmental background developments, vacuous definition, incomplete or even lacking derivation or dating, highly selective etymological data. All these are exemplified in detail below.

Category N (of non-occurrence): 1 item

Strangely, the OED does not recognize an item (to do/pull a brodie) substantiated in three of the sources (Brewer, Funk, and R&S). This omission might be due to the idiom being

essentially an Americanism, which might have been deemed as never having gained ground in Britain; or, alternatively, and

perhaps more probably, it became rejected owing to its scanty record in British print.

Category 0: 1 item

One case emerged to earn the rating 'nought' (as poor as Job's turkey), explained by the OED authorship with a mere note. All critical aspects, starting with definition, were not as informative as one would hope ("a type of poverty"), accompanied with a derivation and background data which were all less than adequate. Furthermore, the dates carried by the first quote in the OED were overdue by some four decades. In other words: on occasion, the OED can also be superficial and uninformative to the extent to justify grounded criticism.

Indeed, such instances, met with unexpectedly frequently, would degrade this great dictionary, without its supremacy in dating, among the mass of mediocrity. In conclusion, this was among the most stupefying observations on "the largest

dictionary in the world".

Category 1: 6 items

To score one point is attributable to various inadequacies:

on the part of smart aleck and clever dick the OED fares

little better than the compilations; in other words, the single point is won by furnishing records of the first instanced use. For the next item (to raise Cain) the OED reaches its lowest yet record by not merely neglecting to define the phrase in an appropriate way, but doing it in such a roundabout fashion as lends itself to mislead the student:

the idiom is simply and bluntly quitted with "to behave in a disturbing way", very much in line with "a type of poverty"

for as poor as Job's turkey. As a result, a not so well-read student arriving from, say, central Africa, or the Indian subcontinent, and hence likely to be unfamiliar with the Bible, would simply become no more knowledgeable of the relevant reference of these idioms in the face of such

inadequate definitions. To complete this picture, two other idioms of this category are treated with similar oversight:

as pleased as Punch is covered with a mere note "an allusive phrase" with its exact meaning being left obscure; likewise, to grin like a Cheshire cat is left without any definition at all.

The most significant drawback in the case of before you could say Jack Robinson is the roundabout treatment of the idiom in the OED: it is aptly defined and duly furnished with datings, but its nonexistent derivation together with the genuinely inadequate background data is worth decreasing the status by a grade to Category 1. To be more precise, the OED does not even touch upon the viable possibilities presented

amply in the source compilations, half of them (4)

elaborating on citations from Dictionary of Vulgar Tongue (1785) by Grose, some on those from The Archaic Dictionary (1846) by Halliwell. These records are totally ignored by the OED.

Category 2: 37 items

What characterizes the representatives of this category

appears to be linked with several elements. In summary of all these various features, one common factor perhaps could

justify the mention; namely the apparent overcaution and unwarranted reserve emerging as a kind of keynote in the way of coverage adopted and maintained by the OED. This is

strikingly illustrated by the examples Black Maria, Soapy Sam, and Uncle Tom. For each of these cases several sources can suggest a perfectly legitimate and viable derivation, which is not even touched upon by the OED as a potentiality.

As noted on some occasions earlier, the OED and the authority it carries would lose little by at least discussing the

evidence strongly advocated by a number of authors of idiom compilations. No less than 13 cases out of the total 37 (roughly a third) are of the type depicted above.

A lesser share among the items in Category 2 for the OED are those furnished with overly vague definitions (in

addition to the two examples (to raise Cain; as poor as Job's

turkey) given in Category 1, exemplified in doubting Thomas, to take the mickey out of someone, and to grin like a

Cheshire cat. There are, in other words, at least five cases (five per cent of the 'Top Hundred') where also the OED has been incapable of an adequate definition, i.e. one that is sufficiently informative the present writer. The remaining 19 cases are those where the OED has, it is true, provided an adequate or even excellent definition, but only rather

unsatisfactory accounts on both derivation and dating in one and the same case.

Category 3: 55 items

Obviously, this category is equally well annotated on all three counts as its equivalent in the sources. On occasion, the OED can excel any of the source compilations by

furnishing an entry that comfortably fulfils each three criteria. Some cases exemplifying this are idioms no. 38 (Colonel Blimp), 46 (Buggins's turn), 50 (Parkinson's law), 60 (to put on the ritz), 71 (mumbo jumbo), 78 (sweet Fanny Adams), 81 (Tom Tiddler's ground), 88 (shipshape and Bristol fashion), and 89 (to talk Billingsgate).

What is nevertheless remarkable in a final overview, is that the OED has thus excelled only in 55 cases out of a hundred against the 73 cases for the compilations. In other words, in a further 17 cases the most comprehensive among the

consulted idiom dictionaries could come up with a better overall coverage than the OED.

In a nutshell, with certain exceptions, the OED

authorship really can produce a well-drafted entry in terms of defining, deriving, and annotating (quoting instances of recorded use and furnishing other relevant and illustrating background data) English idioms containing a proper name. In summary, the OED has its weaknesses - sometimes failing to an astonishing degree - but, equally convincingly, it can also excel over all other reference works in terms of appropriate, adequate, and relevant data.

One of the working presumptions for this Category was that all the idioms having a literary foundation, i.e. being substantiated in print (whether fiction, newspaper articles, documented stage shows, songs, etc.) were included in this category since some form of record of their birth had

survived, in contrast with idioms that had been born

'illiterates', in live conversation, and hence so subsisted until they found a 'literary home' in any of the above

formats. Accordingly, the latter group scored generally one grade less, i.e. Category 2 and downwards, since their very birth was, as a rule, obscure. The few exceptions which would prove the rule also in this context are exemplified by Black Maria, gerrymander, to pull a brodie, all of which apparently came into existence as colloquial phrases, but were soon

accepted into common literary coinage, thus providing a birth

certificate accurate enough within a decade (the first item) or within the actual year (the latter two idioms).

VII.3. FINDINGS ON THE EXPLANATIONS PRESENTED IN THE OED IN