Hi listmates,

thanks for your answers! Definitely the amount of information and views on
the issue is enormous and I think we're nailing down some general points of
agreement on what we could call a "code of good practice".

From the answers so far I'd extract the following points of agreement and
would add some more questions (in the final part). Please, let me know if
there are disagreements, errors, misinterpretations, etc.

1. There are three different translation (by-)products involved in every
translation job:

-a- the translation itself, the document that is the equivalent of the
source text in a different language
-b- TM data, this translation aligned at the sentence level
-c- terminology and phraseology, a subset of the previous that collects
relevant sentence chunks

There is a long tradition about how to go about with the translation itself
so we won't deal with this issue. Terminology and phraseology data tend to
be seen as owned by its producer, although there are different, even
opposite, views on what can be done with this data. In the following, I'll
focus
on TM data (although some parts of the discussion would be applicable to
terminology and phraseology data).

2. There are two ways of using or viewing TM data.

-a- as a productivity tool, i.e. as a way of resuing previous
translations in order to increase consistency and speed
-b- as a a commodity, i.e. as something exchangeable for profit (be it
money or other data)

These two ways of using TM data trigger rather different reactions, even
opposite ones.

3. As a productivity tool, everybody tends to agree that this use of TM data
is OK. Hence, everybody reuses previous translations (that's the reason why
TMs are created and maintained). Besides, there is general agreement that TM
data can be provided, along with the translation job, to contractors. For
example, a translation agency forwards a text for translation to a
free-lancer and provides TM data along with the source text. These are the
uses we have called internal and intermediate.

4. The use of TM data as a commodity, however, opens different
possibilities and reactions.

On the one hand, translation producers (mainly individual contractors) see
this use as a way of broadening their business possibilities: selling a
by-product of their work. And here again we have two different ways of using
this commodity (and two opposite reactions). First, it is OK when TM data
are passed on to the text provider, either as an extra service or included
in the price of the translation. Second, selling this TM data to other
translation producers. This is in general considered wrong. Text providers
argue that selling TM data to third parties would infringe copyright or
confidentiality laws or agreements. Translation agencies tend to agree with
text providers.

5. It's difficult to determine what's legally correct in this context, as it
happens with similar issues in the internet world.

Any information on these issues would greatly enhance our analysis. I'm
going to forward this question to other mailing lists and see what we get.

6. In the absence of general rules, contracts should deal with these issues.

The ATA contract model referes to confidentiality, although not specifically
to TM data. Quote:

<<<7. Confidentiality. All knowledge and information expressly identified by
Client in writing as confidential which Translator acquires during the term
of this Agreement regarding the business and products of Client shall be
maintained in confidentiality by Translator and, except as expressly
authorized by Client in writing, shall not be divulged or published by
Translator and shall not be authorized by Translator to be divulged or
published by others. Confidential information for purposes of this paragraph
shall not include the following:

a. Information which is or becomes available to the general public, provided
the disclosure of such information did not result from a breach by
Translator of this paragraph.>>>

Does anyone know of other contract models that do address the issue more
specifically?

For those interested, there is a section on contracts at
http://www.TransRef.org under section Bussines - Practical issues.

7. Codes of practice should include these issues in order to extend good
practices.

I've checked existing codes of practice and found indirect references to the
issue, like the one in the ITI Code of professional conduct. Quote:

<<< 3.4 Exploitation of Knowledge Acquired
3.4.1 No Member shall derive any gain from confidential information acquired
in the course of his work.
3.4.2 No Member shall accept remuneration from any party in respect of work,
other than as contractually agreed.>>>

Does anyone know of codes of practice that address the issue of TM data in
particular?

For those interested in these codes, some of them are collected at
http://www.TransRef.org under section Newcomers - Deontology.

With my thanks to everyone and best regards,
Antonio