One of the worst nuisances in
my job as a translator for some 10 years now is the constant increase
in ’clever’ glossaries for technical jobs. 

A glossary is fine as such,
provided it deals with technical terms only. It can be a great
help in conveying the client’s preferred choices or internal
trade language, but this requires that whoever is creating the
glossary is sufficiently skilled in linguistics as well as the
technical theme at hand to be able to sift out what belongs in a
glossary and what does not. 

In case the source language
is English, it has to be taken into account that one word in English
may be used as a noun, a verb, an adjective, and an adverb. This is
not necessarily the case in the target language, and most certainly
not in my native language, which is Norwegian. 

Please consider the following
exchange I had with a customer (agency) this summer concerning a
client ”Quality Report” which consisted of an automated
comparison of the translated text against the glossary: 

You
asked me to accept the glossary as is, but that becomes very
difficult as a huge part of their 'errors' are just different verb
declensions, definite/indefinite form of nouns, and similar. I don't
know their proficiency in Norwegian linguistics, but what they're in
fact asking me to do is to disregard a multitude of basic grammar
rules just to make the text fit the glossary. I hesitate strongly to
do that, and I feel it is my duty to make you/them aware that the
text will be severely mangled and quite possibly embarrassing to the
client in the Norwegian market! 

The
problem is that they have erroneously supposed that a term may be
used as both a noun and a verb, like in English, and also that the
glossary may state, for example, the present tense of a verb, while
the context may require imperative tense in the text - totally
altering the meaning! 

An
example of the present/imperative problem would be "Before
accepting an XXX plan, check that the XXX time and the coordinates
for each YYY are reasonable as judged by your experience."
Client Comment: "Glossary Entry: Check => kontrollere"

But "kontrollere" is present tense, not imperative, thereby
losing the notion that you should check the XXX time...

As
for the noun/verb problem, it would be a bit similar to say in
English: "If you own a horse, you may be "horsing"
(riding) around the countryside all day long" Maybe a
far-fetched illustration, but the point should be clear –
sometimes the English original has to be rephrased in Norwegian to
convey the intended meaning – you can't just assume that the
translation of a noun will be exactly the same as a verb. 

This
is the very weakness of using strict glossaries for translation jobs
– either you have to break grammar basics to fit whatever is in
the glossary into the text, or you have to rephrase the sentence to
convey the meaning faithfully. Looking at the Client Report, I'd say
it has not been prepared by a sentient being, most probably it's just
an automated comparison between the text and the glossary with no
consideration of what is suitable in the context. Accordingly, we
cannot accept that all noted ’errors’ actually are
errors.

I
painstakingly went through all 1,239 ’errors’ in their
report and compiled the following statistics (The total word count
for the job was 91,656, so the actual error count wasn’t all
that bad...): 

This clearly shows that a
carelessly prepared glossary adds no quality gain, just a lot of
wasted time and work, both for the client and the translator. 

I would advise anyone
involved in preparing glossaries for translation jobs to use only top
notch translators who are capable of understanding the implications
of choosing which terms should be included in the glossary, as well
as understanding which words may have several meanings. There is no
sense in putting ”time” in the glossary, translated as
hours/minutes, if the usage is ”1st/2nd
time”. This was in fact done in the case story above. 

Above all, do not

think that you don’t need a highly skilled / expensive
translator to prepare the glossary, as if ”It’s just
single words, anyway”... 

In fact, translating the
glossary is far more demanding than working with the actual text due
to the lack of context in a glossary, requiring the translator to
know what the component/unit is and what it does, without any
support from the context.  

In highly specialized,
technical texts, this is not as obvious as one might think – I
shall not soon forget the ”Pitman” in the manual of a
rock crushing plant some 7 years ago. It was listed in a maintenance
schedule, and for the life of me, I couldn't understand why this
mining worker should be monitored/tested for fatigue every 100
hours... 

Well, it turned out to be the
”piston rod” driving the actual crushing jaw on the
machine, not so redily deductable from a ’dry’ text with
no illustrations, I’ll say!

The following procedure will
secure that your (the client’s) prepared glossary will actually
add to the quality, not just create wasted manhours in the project:

1.  
Have your Engineering/Construction Department create a list of
components/units/parts that are unique to the equipment/technology,
and ask them to omit standard parts, such as valves, pipes, brackets,
plates, tubes, fittings, and handles, unless they are designated in a
special, preferred way in your manuals

2.  
Hire a high quality translator (agency) to translate the list,
making sure ( by references, samples, etc.) the candidate is
sufficiently experienced in the technical field at hand. Do
not pinch pennies on this step!

3.  
Have the translation validated by your native
distributor/affiliate company, securing that all translations are in
line with the preferred terminology/trade language at their site

4.  

Have the validated translation scrutinized by a skilled
linguist in the target language, sifting out possible grammar errors,
typos, or overly colloquial language

5.  
Now you are ready to hand over the glossary to the
translator/agency and start the actual translation project

Keep in mind that your
glossary will be your foundation for the whole project, meaning that
errors/ambiguities in the glossary will be propagated throughout your
manuals, so this is the place and time to spend the bulk of your QA
funds!

Løfallstrand,
2008.11.22

Roald Toskedal

- Partner &
QA Manager -

Norwegiantrans
ANS