The very first translation agency that hired me, forced me using Trados GmbH Workbench 5 and after useless bargaining (they were entirely resolute) I had to accept it against the promise of a lot of work (that luckily went)
These years, the chance to find Trados memories was a daydream, or maybe I was simply less able than today to drive out them, but the situation changed less than a decade after
In 2010 after my entry in the database of one of the top 10 global language providers, I acquired a lot of huge and (apparently) well managed Trados memories, and at the same time I was able to find others for free
For example, in 2010 and due to a lucky break, with €60 I was able to download all TAUS English to Italian corpora uploading the DGT memory (they had it already, but quite oddly they called it differently ;-)
This finding got me full of high spirits, and the same effect did driving out the DGT memory, then the Opus TM, and again the EURLex database (so easy to align with Logiterm Pro in those days, now with NOVA Text Aligner)
But very soon my excitement was dampened by the low quality of (too) many of these resources: incoherences, grammar and spelling mistakes, and even worst a lot of awful technical terminology blunders even in more (apparently) authoritative corpora as EU Directives, for example, pushed me back to CD/DVD-ROM dictionaries
But as some of those corpora, for example EU Directives or certain EMA/MedDRA files, are mandatory in certain medical fields, I found myself in front of a dilemma: whose resource believe to, more, CD/DVD-ROM dictionaries, laws, EMA files, other?
After a lot of reasoning, I conceived a sort of "command of chain" for the medical terminology pipeline (but one can use it in any other field, doing the needed changes) that I want to share now
The rule is that the main context shows the Boss, so usually there is one Boss and as many Deputies as secondary contexts exist
For example, imagine to receive for translation/proofreading a PIL, SPC + label of a human drug: In this case the Boss is the last version (check it here and there) EMA QRD template WC500004369 because the EMA application can be rejected if the writing is wrong (or the layout is not accurate enough, BTW)
Every sentence in that file has an exact translation in every language and you MUST follow it, even if the translation seems wrong or odd!
For example, in the SPC (SUMMARY OF PRODUCT CHARACTERISTICS) at point 4 you find:
CLINICAL PARTICULARS and in Italian you must write INFORMAZIONI CLINICHE and nothing else
So even if PARTICULARS can be translated with synonyms as PARTICOLARI, DETTAGLI or DATI, you cannot put PARTICOLARI CLINICI, DETTAGLI CLINICI, or DATI CLINICI, you MUST use INFORMAZIONI CLINICHE in any case!
And if the source says (wrongly) 4. CLINIC DETAIL, you must put your translation as it would have been correct i.e. CLINICAL PARTICULARS
It means that:
4. CLINICAL PARTICULARS = 4. INFORMAZIONI CLINICHE
4. CLINIC DETAIL = 4. INFORMAZIONI CLINICHE
(to be continued)