Some recent OED releases: a tale of two treats

As each fascicle of the original OED was published, it became customary for it to be prefaced by notes drawing attention to some of the words included which were of particular interest, especially as far as etymology was concerned. For example, when the range of words published in OED Online in December 2000, marciaton-massymore, appeared in print in 1905, particular mention was made of the following words as being 'interesting on account either of etymology (on which in many instances new light has been thrown) or of sense-development': marmalade, marquis, marriage, marry, marshal, martinet, martingale, martlet, marvel, marzipan, mask sb.³, mason, masquerade, mass sb.¹.

A striking illustration of the extent of the revision work we are now carrying out is the fact that all of these entries, selected by editor Henry Bradley as 'highlights' of the range, now contain so much new material of interest 'on account either of etymology...or of sense-development'. Much of the new information comes directly from our own primary research, examining the historical record of English; but much also depends on the new light cast by recent scholarship on the history of other languages, as we can see by looking at just one example from this list.

The OED's coverage of the word marzipan has undergone profound change. In the First Edition of the OED, if you looked under Marzipan, you would find only the cross-reference 'Marzepa(i)ne, Marzipan: see MARCHPANE'. The entry for this word carried a double headword 'Marchpane, marzipan', and all the different spellings - marchpane, marzipan, and over a dozen others - were all placed together in a single entry, with only the general comment in the etymology that 'the Eng[lish] forms [of the word] come from various continental sources'. In fact the spelling marchpane was predominant throughout the nineteenth century (and indeed long before that), so the decision to place the entry under this heading must have seemed reasonable. However, in the course of the twentieth century the spelling marzipan, and the three-syllabled pronunciation which goes with it, have almost completely replaced the older marchpane: certainly the latter is now rarely found outside historical contexts.

Although marchpane and marzipan both derive ultimately from Italian marzapane, they have a claim to being distinct words, and in the revised OED they are treated as such: the entry for marchpane was published online in September 2000, and marzipan appeared in the next batch, three months later. The entry for marchpane now offers an explanation of how this form of the word - first found in English in 1516, spelled 'march payne' - came to be so different from its Italian etymon. The first element, as well as showing an English sound change, may have been influenced by March (the month, named after the god Mars). The etymology cites parallel forms in Latin dating from around 1500 (martiapanes and panis marcius), both suggesting that the word was thought to be linked to March or to Mars in some way. And the second element, -pane, seems to have been assumed to be the same word as Latin panis or French pain. Certainly this 'more Anglicized' form of the word seems to have been in more general use in English than the 'foreign-looking' marzipan: we do know of spellings similar to the latter from as early as 1542 ('marzepaines', in an English translation of Erasmus), but they remained rare until the 19th century.

What, then, is the ultimate origin of marzipan (and its cousin marchpane)? The original OED entry comments that 'Its etymology is obscure', and does no more than mention one scholar as having 'ingeniously' suggested a link with 'Arabic mauthabn "a king that sits still"'. Once again, recent scholarship allows the new OED entry to put forward a new derivation: in this case Italian philologists have furnished the basis for a link with - remarkably - the Far East. In Myanmar (Burma) there is a port, near the town of Moulmein, called Martaban, which was famous for the glazed jars which it exported to the West, often containing preserves and sweetmeats. Delicacies are often associated with the containers in which they are traditionally imported (ginger being an obvious example); it seems plausible enough that a name associated with a special container should transfer its association to the thing contained.

Plausibility would not, however, be enough were it not for a curious aspect of the words which correspond to marzipan in some of the other European languages: Italian marzapane, Spanish mazapán, French massepain. In each case the relevant word once also had another meaning, denoting various kinds of container - a casket in 15th-century French and 14th-century Spanish (specifically for confectionery in the case of French), and a container of a certain capacity in Venetian documents in the 13th. And then there is also the fact that Martaban is still known for its pottery: the same batch of recently published OED entries which contains marzipan also contains an entry for Martaban jar (sometimes simply Martaban), this being a kind of large glazed earthenware jar. (The same jars have also arrived in English via Afrikaans: the ships of the Dutch East India Company carried them to South Africa, where even English speakers came to call them Martevaans. By the same exacting criteria that separated marchpane and marzipan, we distinguish Martaban (jar) from Martevaan - the latter has its own entry in the OED, now published for the first time.) Thus in these entries, as in almost every entry published online to date, we can demonstrate that we have learnt, and can pass on, more information about 'etymology...or...sense-development' than ever before.