User:DCDuring/Review of OED3 from ADS-l

A quick review of the new OED.com Inbox ADS Daphne Preston-Kendal dpk@nonceword.org via listserv.uga.edu 7:12 AM (5 hours ago) to ADS-L Today when I visited OED.com at home I saw that the new redesign has been launched. Or at least semi-launched: when I connected to my university’s VPN, which I need in order to actually see entries, the old design came up. So I am only able to review the selection of entries which are available without logging in (today’s ‘word of the day’ and the ‘recently added’ and ‘recently updated’ entries), and I can’t test things like the search features yet. Nor do I know what ‘OED2+’ entries (those not yet fully revised for OED3) look like, since all the free entries are updated OED3 ones.

My impression is that the changes made are a mixed bag in terms of whether they are improvements or disimprovements. On the one hand, this is certainly the first OED.com which comes close to making reasonable use of the information design capabilities of an interactive medium, rather than exactly replicating the layout of the printed dictionary page. On the other, these capabilities are arguably not always used well. I think I will need more experience before forming a solid conclusion.

It is now much easier to find the sense you’re interested in within an entry. All quotations except the first and last in a paragraph are now hidden by default, though there is still a button to unfold all of them in the entry in one go. This makes it much quicker to scroll through a long entry scanning for keywords in the definition text related to the sense one is looking for. The left hand column also features a ‘table of contents’ where the incipit of each sense’s can be read – clicking the text scrolls the page to the sense.

It is also now easier to understand where one is in the sense hierarchy of a more complex entry, because each sense number is given in full next to each individual sense (i.e. the sequence is now shown as 1.a. 1.b. 1.c. instead of 1.a. b. c.). This could be further improved if some kind of highlighting were applied to the table of contents in the left-hand column to reflect the current page scroll position. (This is used in the new design of Wikipedia.)

Among changes I will have to reflect more on before coming to a conclusion about their utility is the decision to move certain information to different ‘sub-pages’ of the entry by default, which are accessed by clicking on different ‘tabs’ at the top of the entry. Essentially all information which was previously at the top of an entry before the definitions now each has its own sub page, including the etymology, pronunciation, variant and historical forms, and frequency. A new ‘factsheet’ tab of dubious utility summarizes all this information; more positively, there is now a ‘compounds and derived words’ tab which a similar function to a ‘what links here’ feature on a wiki: it lists entries whose etymologies link back to the entry you’re currently looking at. (I intended to add such a feature to GDoS Online but never got around to it before leaving the project in 2018.)

In general, I think hiding information behind an extra click – especially when that click itself causes other information that was already there to be hidden – needs a strong justification. I’m not sure everything that has been put behind an extra click is strongly justified in this case. Making the default view begin at the definitions certainly addresses previous philosophical criticisms of the classic OED model where the earliest sense directly follows the etymology; also, since OED3 etymologies can be fairly discursive, it decreases the amount of page scrolling needed to get to the definitions, if that is what one is primarily interested in. However, it emphasizes the strangeness of one OED3 editorial decision dating from the start of the new edition: the fact that etymological information relating to particular senses is not placed inline with each relevant sense, but instead listed in notes at the top of the entry in the main etymology section. In the entry ‘docket, noun 1’, one is thus in the position of being on the etymology page where one is told that ‘Sense 1 may show a different word’ without any indication of what sense 1 is – in the reverse situation, a small-type note attached to the definition at least refers one to the etymology on this subject, but on the version of the dictionary on the old website this was not done consistently, as far as I recall.

The decision to move the pronunciation to its own page is among the most bizarre, since there is so little information on that page, and it could easily have been fit into the entry heading. On the other hand, the OED is not primarily a pronunciation dictionary and I doubt anyone is using it as such. Since OED2 adopted the IPA, one has to look at the variant forms and have a strong sense for how the historical phonology and orthography of English intermesh to get any accurate sense of the historical phonological form of a word in any case.

One can also disable the ‘tabbed view’ with a switch in the upper right, which puts things all in one big page to scroll through, but still in reverse order compared to the old OED: you have definitions, then etymology, then pronunciation, then forms, then frequency. You can jump between the sections easily because they appear as new, higher-level rows in the table of contents. This preference appears not to persist between looking at entries, so if you disable tabbed view in one entry, the next one you look at will bring it back.

In other, more-or-less inconsequential design changes: Small capitals are now gone entirely, from both cross-references and quotation author names. Quotations are now given in 2DWB style (date – text – citation) rather than classic OED form (date – citation – text). The names of languages in the etymology are now given in bold type for some reason, which I can imagine will be quite distracting in entries where lists of Germanic and Indo-European cognates are given.

Some more abbreviations have been expanded: ‘tr.’ in quotations is now given as ‘translation of’. An attempt has been made to show the titles of cited works ‘in full’ but prepositions and articles are still often missing. (Examples from today’s word of the day: ‘Narrative of Troubles with Indians New-England’; ‘Investigation Causes & Consequences of Present Contest with Holland’).

None of my concerns about the long-term sustainability of digital OED publication (which I raised at ICHLL12 – the proceedings volume is in press) have been addressed in the new version. OED2 versions of entries are still apparently available to view (‘details’ link in the upper left corner, then ‘View in Second Edition’), but what they now look like, I don’t know, because that also requires logging in. It’s my understanding that there is still no way to run a search in the OED2 text. URLs to entries now look different but I’m not sure if old ones still work and redirect, or if all those older links are now broken.

Once I see that my university has been granted access to the new design, I’ll try out some of the things I wasn’t able to use for this review.

Daphne Preston-Kendal

The American Dialect Society - http://www.americandialect.org

Jesse Sheidlower via listserv.uga.edu 10:27 AM (1 hour ago) to ADS-L Thanks for this extensive writeup, Daphne.

While I have no formal connection to OED, I did have a chance to review the new site pre-launch, and to discuss many of the new features. I can say that OED staff are aware of pretty much all of these concerns; I think I can also say that there are, in many cases, differences of opinion between lexicographers and others about many aspects of the site.

The access issues are probably caused by the length of time it is taking for the switchover to propagate across the globe; this should be fixed very soon.

Some specific comments:

The tabbed view (which seems to be widely disliked by most lexicographers) can be turned off by setting up a personal account and configuring "Display preferences" to always default to the non-tabbed view.

All previous entry URLs will redirect correctly. (At least, they will when the access issues are fixed.)

The "factsheet", which I personally have an extremely low opinion of, is mainly intended for non-registered users of the site; if you do have OED access, you should never see it unless you choose to do so. I suppose this makes some kind of sense, though I do think that for it to be at all useful for non-registered users, it should at least include a brief gloss of the word in question.