A Pragmatic Eye

On Friday 18th Jan, I attended a most informative talk at Europe House, Smith Square, Olde London Towne on the subject of terminology and IATE, given by the knowledgeable and engaging Timothy Cooper, a senior terminologist at the EU.

The theory

He kicked off with an overview of the message he attempts to drum into all those on the 9-month (!) probation at the EU Translation Service (8 months of translation then a month working on terminology), to wit terminology is not about terms, it’s about concepts. Apparently, this is quite incendiary stuff as far as academics are concerned, although Tim didn’t say what the opposing view was. Apparently too, some probationers also have difficulty grasping it. Given that the echoes of lectures given by academics are probably still ringing in their ears (behind which traces of wetness no doubt remain) perhaps this should be no surprise (it seems the route in EU translation is quite often straight from study into the EU).

I doubt any readers of this would need Cooper’s key message to be exactly hammered home with a mallet, the underlying idea being (and I quote, more or less) that the concept is the mental abstraction of an object (tangible or intangible) which has a definition to distinguish it from other concepts. The terms emerge once you’ve got a clear picture of what you’re talking about. I’m sure we all regularly come up against terms where we need to get a clear definition of one or several possible translation options before we can be sure the source term (for which we have also determined a definition) and our selected target term really are one and the same concept. That’s the basic idea, illustrated by a simple example as Tim worked on the translation of the term “coeur” and got the results “heart” and “core” for anatomy and nuclear power contexts respectively, via concepts and definitions.

So, to IATE itself. First point to note is that it stands for Inter-Active Terminology for Europe. The second point to note is that for those of us outside the glorious institutions of Europe, a more accurate abbreviation would be TE. It is not interactive for us. It’s interactive for them; the version we see is updated monthly-ish and is purely read-only.

It covers all EU institutions – commission, parliament, council and so on – in all fields of activity, but is restricted to LSP (language for special purposes).

The general idea is that structurally, it kinda sorta reflects the principle outlined above. There is a notional hierarchy topped by “language independent information”, such as the domain (of which more in a sec), then below that information in various languages, such as the definition, references and notes, then at the third level sit the various terms in those languages. Some languages are better represented than others – typically the longer a country where the language is spoken has been an EU member, the more terms there will be. (Internally, the system holds terms in 100 or so languages, but only official EU languages are displayed to the masses.)

It should be noted that while Tim Cooper advocates the “substitution” approach to definitions (whereby the definition is a phrase that could simply replace the term in any context and it would make grammatical and logical sense), not all of his colleagues in other countries (countries now populate the database more or less independently) share his clarity of vision. Do a search on “financial service” (note that IATE doesn’t like plurals) and compare the English definition with the Lithuanian one.

The “domain” field was mentioned mainly to warn us not to pay it any heed. It is ostensibly based on the Eurovoc domain system. Frankly, I was reminded of one those ancient Greek fellows (or possibly a Victorian clergyman) who attempted to categorise all of human knowledge. It’s bound to be fairly arbitrary, and as Tim Cooper freely admitted (having described Eurovoc’s domains as “nonsensical”), in IATE’s case, there are too many errors even in clear-cut cases (he cited land transport terms being categorised under air transport or vice versa). He recommends not using it for searches unless your first search gets an overwhelming number of hits.

The practice

Which brings us neatly to searches. (As an aside, note there is an option to save your search preferences.) The first and key piece of advice is not to search for a specific pair; use the “Any” checkbox to the right of the list of target languages. Note the preferred spelling is UK English spelling (e.g. -ise not –ize), but alternative spellings and indeed plain ol’ mistakes might still get hits if the terminologist has added such entries to a device known as a lookup form (do a search on “accomodation” with one ‘m’ to see this in action – hits with the correct spelling are found, as is at least one rather embarrassing entry spelt incorrectly).

The list of hits is in list form, with a “full entry” link to click to show all the details. The list shows a reliability rating measured in stars. Ignore it. Again, different countries have different approaches. You need to follow up references given in the full entry screen to be sure how reliable an entry really is. Sometimes, the list may show terms that are marked as “Obsolete” or “Deprecated” – those are best avoided (unless, perhaps, you’re translating something from when those terms were current).

The joy of the “full entry” screen is the references and links elsewhere. This is especially useful if a link points to EurLex, these being links that you might see given in any language but not necessarily all of them (even if it would be relevant), because once you’re in EurLex, you can switch to any other EU language. Hence the advice not to search just for your language pair – you are potentially losing valuable links to other multilingual information. And that, in Tim Cooper’s view, is the value of IATE. In itself, the content leaves something to be desired (I paraphrase), but it can lead you in the right direction (much like Wikipedia, a comparison he made himself).

On the “full entry” screen, you might see cross references to other IATE terms (IATE:nnnnn). These should be hyperlinks. They are internally. They might be for us, for one day….

The original data was imported from various sources and because of problems with mapping equivalent fields, there was some inconsistency which remains in the older entries (hence dodgy definitions, unreliable reliability ratings, etc.). These days, data is added manually (if I understood correctly) by the different translation services in different countries, and while there is more consistency in approach, it is not completely consistent (how they decide on the reliability rating, how much effort they put into lookup forms, the definition of a “definition”, etc.). But as a guide, the more recent an entry is, the more reliable it should be.

Summary search tips:
– Don’t use plurals
– Ignore domains
– Search on all languages, not a pair
– The “reliability” star rating is not to be relied upon.

As you can perhaps tell, I found the whole thing very interesting and useful. Before attending, I would have described myself as a former user of IATE. I suspect that is about to change.

NB: In case anyone is wondering whether I’m taking an awful liberty posting a large amount (if summarised) of the actual content of Tim Cooper’s talk online, it was free of charge (discounting the absolute battering my Oyster card took – that daily cap TfL claim to apply? Be suspicious…). The event was open to anyone, and I’m sure their aim is just to spread the word. So I’m just, y’know, spreading the word!