Reading Time: 2 minutes

Some dictionaries are designed to explain what words mean. Some dictionaries, like the Oxford English Dictionary, aim to record changes in the the meanings of words over time, and attempt to identify the earliest use of each word (or if they have multiple meanings, of each meaning).

Geoffrey Hill, in his review of the Second Edition of the Oxford English Dictionary, published in the TLS of April 21 1989 (and reprinted May 10, 2019), does not appear to understand what the OED aims to do. His review uses the poetry of Gerard Manley Hopkins to justify his claim that the compilers “failed to register the metamorphic power of Hopkins’s content”, and he gives for example the word “disremembering”, which Hopkins uses in the poem “Spelt from Sibyl’s Leaves”:


Disremembering, dismembering all now.

The word “disremember” is certainly unusual. The OED records it being used in a few other places during the 19th century, with the meaning “failing to remember”. For Hill, the meaning in this poem is clear. It is not “failing to remember” but “dismembering the memory”.

No lexicographer, even in the Oxford English Dictionary, would attempt to capture metaphorical use of language in poetry. It is perfectly possible that Hopkins meant both “fail to remember” and, as Hill claims, “dismembering the memory”, but that doesn’t mean that lexicographers should dutifully record a one-time usage that is not replicated by other users elsewhere. Very strangely, Hill insists on his own meaning and denies the common, and recorded, meaning of the term.

Of course, the digitisation of texts means that the creation of a corpus-based dictionary such as the OED is in some ways easier. If all published text is available in digital form, the OED may be able to draw on not four or five citations, but hundreds. That doesn’t mean that each citation represents a valid example for the lexicographer to try to define, and certainly not use in a metaphoric context.

For Hill, the OED “patently fails to register the metamorphic power of Hopkins’s context”. It’s just not what dictionaries do. I would argue in this case that the dictionary has done exactly what it should do, and no more.