Mobile menu

Pages in topic:   [1 2] >
Are we cheating ourselves by using analysis tools?
Thread poster: Heinrich Pesch

Heinrich Pesch  Identity Verified
Finland
Local time: 15:38
Member (2003)
Finnish to German
+ ...
Feb 18, 2005

How come that Trados and Wordfast, when using their analyse-funktions, leave numbers out from the beginning?
We probably all agree, that a sentence with one or more number in it requires more work than a sentence with only text. Though of course the number does not need to be translated, it needs to be preserved, which means either I have to interrupt the typing, copy the number from the source segment into the target segment at its proper place, or that I copy the source completely into the source and start to edit it piece by piece arount the number(s). Either way I spend more time on this segment than on a segment of plain text of comparable length.
And get paid even less.

How come, that we agree to this? Are we nuts?
What do you think?

Regards

Heinrich


Direct link Reply with quote
 
xxxBrandis
Local time: 14:38
English to German
+ ...
difficult to understand Feb 18, 2005

Hello Heinrich, Do you mean, that you are faster when you do not use a CAT-Tool? Wouldn´t you be putting up some kind of glossary of the terms, while translating without the help of any translating. How would you be counting the text, with all the repeats, part words etc., I am trying to grasp your idea of translating with the help of any CAT. One method I know is hand finished, where every repeat would have probably a different translation.
Rgds, Brandis


Direct link Reply with quote
 

Mihail M Mateev
Bulgaria
Local time: 15:38
Member
English to Bulgarian
+ ...
Wordfast uses short-keys combo's to copy/paste the nimbers Feb 18, 2005

Heinrich Pesch wrote:

How come that Trados and Wordfast, when using their analyse-funktions, leave numbers out from the beginning?
We probably all agree, that a sentence with one or more number in it requires more work than a sentence with only text. Though of course the number does not need to be translated, it needs to be preserved, which means either I have to interrupt the typing, copy the number from the source segment into the target segment at its proper place, or that I copy the source completely into the source and start to edit it piece by piece arount the number(s). Either way I spend more time on this segment than on a segment of plain text of comparable length.
And get paid even less.

How come, that we agree to this? Are we nuts?
What do you think?

Regards

Heinrich


... so I am not agree with you.


Direct link Reply with quote
 

Heinrich Pesch  Identity Verified
Finland
Local time: 15:38
Member (2003)
Finnish to German
+ ...
TOPIC STARTER
Well, here an example Feb 18, 2005

Let's take this piece of text:

"Press and hold [+] for approx. 3 seconds while "3, 2, 1" appears in the display. The instrument’s usable life is now
started. The gas to be measured will be shown. After 10 seconds the display will turn off or if [+] is pressed again,
"d" will be shown. After 10 seconds the display will turn off or after another press the remaining time in days will be shown.
After 10 seconds the display will turn off."

When I use Word's statistics function, I get: 402 characters, 77 words.
With Wordfast analyse I get: 397 characters, 76 words

With Trados WB 6.5 I get: 329 characters, 76 words (so if I go with the Trados analysis I would charge the customer 18 % less than according to Word's statistics).

Now I understand why agencies prefer Trados over Wordfast.

Is that what we translators want? Why do we go with it?


Direct link Reply with quote
 

Christine Andersen  Identity Verified
Denmark
Local time: 14:38
Member (2003)
Danish to English
+ ...
I do NOT work faster with Trados Feb 18, 2005

My experience is that I do NOT work faster with Trados. Terminology is more consistent, yes, but it does not save time, or if it does, it costs extra effort at the proofing stage.

This does have a lot to do with the kinds of texts I translate - they are meant to be read, so often where there are repeats, I rephrase the sentence at the proofing stage anyway, because it repeats too much and the text sounds 'dead'.

Simply replacing 'however' with nevertheless / although / even if

alternating between things like 'besides', 'additionally' etc.
or: therefore / this means that / in other words

... does wonders.

Other times I reverse the order of main and subordinate clauses, which Trados cannot do.
Sentences that sound fine as
1
2
3
the first time they come up may sound very awkward between other sentences later in the text as
17
2
21
3
or when they appear as 90% matches...

Trados cannot adjust the rhythm of the sentence to vary what I suppose would be 40% or 50% matches, or where similar, but not identical terms come close together and need special attention to distinguish them.
................

On other occasions, when in practice only certain portions of the text are being updated, it is good to rattle through unaltered sections with Trados, but it was almost as easy in the old days with 'compare documents' where it worked, and 'cut and paste'.

So in short, Heinrich, yes, we are cheating ourselves when we use the analyses!


Direct link Reply with quote
 
xxxMarc P  Identity Verified
Local time: 14:38
German to English
+ ...
More fundamental issues Feb 18, 2005

There are two, more fundamental issues here.

The first is that volume is a very crude way of charging for translation services.

If you go into a bookstore, you'll generally find that the bigger and heavier books are also the most expensive ones. There is a correlation between the weight and the price of books. I have heard of 2nd-hand bookstores selling books no one wants to read by the kilo - and the locals filling their cars with them and using them as heating fuel. When I bought a pile of books at my local flea-market last year, the seller put the pile against a ruler and charged me by the centimetre. But no one would expect a normal bookstore to sell new books by the kilo. The price is determined by two factors: the accumulated costs of the value chain, and the perceived value to the purchaser.

Volume is a convenient way of selling translation services, but it can only be a very crude basis. When a difficult text can take three or four times as long to translate as an easier text, it's silly to be knocking bits off the price here or there for numbers, or because spaces are easier to type in, or because the words are shorter, or for whatever other reason the bean-counters who understand nothing about translation dream up. (Strangely, the bean-counters never find reasons for adding to the price of a translation.)

The second issue is that the mode of charging for a professional service is the professional's prerogative. I have no objection to other translators not charging for numbers (or for spaces, or for any words beginning with "w", or for whatever they like) - that's up to them. I am not going to be told by other parties, though, in particular not by software vendors, what I should and shouldn't charge for. I'm not charging for words, I'm charging for the provision of a translation service. It's just as easy to find reasons for bumping up the price as it is for lowering it: for instance, a financial translator might encounter one page of technical data within a financial text which requires the purchase of an expensive technical dictionary. Or some aspect of the text might necessitate a long-distance phone call to the customer. Why not put that on the bill? Surprised? A solicitor would, believe me. I once asked a solicitor what he would charge to prevent a landlord from evicting me. His reply: "600 Deutschmarks for writing to the landlord, plus the odd photocopy, 800 Deutschmarks in total". Perhaps one day we'll see "Solicitors Workbench", the tool that churns out standard solicitors' letters to landlords for a standard end price, plus three cents per photocopy. Totally ridiculous, of course - I wasn't paying for letter, but for a professional's expertise.

Analysis functions in CAT tools can be very useful, but how they are used should ultimately be at the translator's discretion. Some translators charge for all numbers; some charge for numbers within text but not for large volumes of numerical data, for instance in tables; some don't charge for numbers at all. There is no consensus. It might be nice to find a consensus (personally I think not, for the reasons given above), but the notion of a software vendor deciding the charging arrangement is as silly as a screwdriver manufacturer saying how electrician's fees should be calculated.

Marc


Direct link Reply with quote
 

Heinrich Pesch  Identity Verified
Finland
Local time: 15:38
Member (2003)
Finnish to German
+ ...
TOPIC STARTER
Thank you Marc! Feb 18, 2005

At last someone who grasps my intention:
Marc wrote:

"Analysis functions in CAT tools can be very useful, but how they are used should ultimately be at the translator's discretion. Some translators charge for all numbers; some charge for numbers within text but not for large volumes of numerical data, for instance in tables; some don't charge for numbers at all. There is no consensus. It might be nice to find a consensus (personally I think not, for the reasons given above), but the notion of a software vendor deciding the charging arrangement is as silly as a screwdriver manufacturer saying how electrician's fees should be calculated."

I believe we as a community should decide, which analysis tool is exeptable. We need a word-wide organisation that tests software for use in translation and awards some kind of certificate, and we freelance translators should use only software that has got this certificate. As a community we are strong, if we only want.

Regards
Heinrich


Direct link Reply with quote
 

Marijke Singer  Identity Verified
United Kingdom
Local time: 13:38
Dutch to English
+ ...
Spot on Marc! Feb 18, 2005

MarcPrior wrote:

Analysis functions in CAT tools can be very useful, but how they are used should ultimately be at the translator's discretion. Some translators charge for all numbers; some charge for numbers within text but not for large volumes of numerical data, for instance in tables; some don't charge for numbers at all. There is no consensus. It might be nice to find a consensus (personally I think not, for the reasons given above), but the notion of a software vendor deciding the charging arrangement is as silly as a screwdriver manufacturer saying how electrician's fees should be calculated.

Marc



Amen to that!

Another issue which has been touched on but not fully explored is that we all specialise in different fields. I mainly translate technical documentation and I am a lot faster with Trados mainly because you have to be repetitive. The last manual I translated used the phrase "Unscrew the bolts ..." 100 times. It would be strange if I suddenly started to write "Loosen the bolts ...". There were variations: "Unscrew the 3 bolts ..." and "Unscrew the 2 bolts ...". Now I know I will not be paid for translating 3 and 2 but since I mainly translate for returning customers, I see it as providing added value which will ensure I keep my customers. The only marketing expenses I have are my subscriptions to this website and a few others. Therefore, I do not mind not charging for the numbers. My point is that it is all very relative. I might lose out by not charging for numbers but I gain in other areas.


Direct link Reply with quote
 

Ralf Lemster  Identity Verified
Germany
Local time: 14:38
English to German
+ ...
Sovereign, individual business decision Feb 18, 2005

Hi Heinrich,
You have raised an important issue, but here I disagree with you:

I believe we as a community should decide, which analysis tool is exeptable.

Determining prices for services rendered is a service provider's sovereign decision. You could say that smart pricing (ensuring that your service is both competitive and profitable) is one of the core elements of a successful business.

That is why - as Marc pointed out - it's the service provider's prerogative to set a price. Metrics (such as word counts, match analyses, etc.) can be useful to find a common basis for pricing, but in the end we need to focus on the revenue per hour (day) worked. This is a decision every person running a business has to make for him/herself - I would not have that decision taken away from me, whether by a customer or an association.

Best regards,
Ralf


Direct link Reply with quote
 

Heinrich Pesch  Identity Verified
Finland
Local time: 15:38
Member (2003)
Finnish to German
+ ...
TOPIC STARTER
That's up to individual believes, indeed Feb 18, 2005

I respect your liberal views, Ralf, in principal.
But what business men are we freelance translators? The only difference to traditional swetshop industries is, that we sit alone, some swetting, some freezing.
Yes, we can walk away, decline work, stick with good customers. That's what we all do.

But who has given a software firm the right to provide in-built discounts to our customers? They did not ask our opinion, but cleverly offered our employers an advantage.
And because we have no organisation that defends our rights this just happened.

Who's afraid of standardisation, if it serves our interests?

Regards
Heinrich


Direct link Reply with quote
 

Mark Xiang
Local time: 20:38
English to Chinese
+ ...
Heinrich Pesch, ABSOLUTELY AGREE WITH YOU!!! Feb 18, 2005

This is the very subject I was thinking to raise!

Just a few days before, an agency used Trados to cut its 3000-word project to a 1500-word job! I declined the job. I do not use Trados. If there is really large quatity of repetition, I will negotiate it with my client and we will agree upon a word number. If my client want to count out the repetition by using Trados, I will double my word rate.

In my opinion, Trados is not for freelancers. It is for AGENCIES!!! And many peers in my pair say it does not help much.

Mark


Direct link Reply with quote
 

Ralf Lemster  Identity Verified
Germany
Local time: 14:38
English to German
+ ...
Running your business vs. having someone else running it for you Feb 18, 2005

Hi again,
But what business men are we freelance translators? The only difference to traditional swetshop industries is, that we sit alone, some swetting, some freezing.

But isn't that up to everyone?

Yes, we can walk away, decline work, stick with good customers. That's what we all do.

I'd say that this is what we all should do, but I see only a minority really implement that view. Self-employed translators need to have a firm idea about their costs and target returns, yet repeated questions in these forums about "what they should charge" and "what sort of discount they can accept" clearly indicates that this is not the case.

But who has given a software firm the right to provide in-built discounts to our customers?

Nobody. And, in fact, that's impossible, as software firms are not a party to the business relationship between yourself and your customers.

They did not ask our opinion, but cleverly offered our employers an advantage.

"Employers"? Self-employed translators have customers - that's a very important distinction, as it says a lot about the stance taken vis-à-vis the same.

Best regards,
Ralf


Direct link Reply with quote
 

Heinrich Pesch  Identity Verified
Finland
Local time: 15:38
Member (2003)
Finnish to German
+ ...
TOPIC STARTER
Thats already another discussion Feb 18, 2005

"Employers"? Self-employed translators have customers - that's a very important distinction, as it says a lot about the stance taken vis-à-vis the same."

The distinction is very much blurred nowadays. We do not move and work in employer-provided facilities, as do freelance actors, part-time journalists etc. But otherwise we are plain workers, selling our time to those who are paying.

I'm sorry I raised the question of a worldwide organisation, but I feel the question of unfair analysis software should be adressed in detail.



Regards

Heinrich


Direct link Reply with quote
 
xxxMarc P  Identity Verified
Local time: 14:38
German to English
+ ...
Standards Feb 18, 2005

Heinrich,

I do quite a lot of translation work on standards policy, so I know something about it. A couple of points:

Standards can be very useful. The fact that the "Normzeile" is defined in Germany is very helpful; German customers hearing a price per line who don't know how translations are charged for and who have visions of the translator changing the font to 24 point and the margins to three inches are immediately at ease when they hear that the line is standardized. "Standardized" isn't strictly speaking correct, but there is a general consensus that a line consists of 55 keystrokes, that this includes spaces, etc. It isn't a standard in the strict sense of the word but it has found its way into the legislation covering court translations and could quite easily be defended in court.

I can see a case for taking such a standard further and defining in it whether, for example, numbers should be charged for (at all, or only in large bodies of data, or at a reduced rate in large bodies of data to cover the handling, coversion of commas to decimal points, etc. etc.). A more relevant issue in my case is bibliographies. For the most part, they are left unchanged - but they do still require some work. A case could be made for charging a reduced rate for bibliographies because of the lower work they ential.

I'm in two minds about this because, as I say, quantifying the cost of a translation isn't an exact science and the more we pretend that it is, the more we delude customers into thinking that translation is merely the transposition of words. I could make a good case for granting a discount of 50% for bibliographies. But I could make an equally good case for charging a flat fee of 30 euros for every word in the text that isn't in the five most popular general technical dictonaries and the two most popular subject-specific technical dictionaries, on the basis that this metric would reflect the level of specialism of the text. Apart from making pricing even more complicated than it already is, though, this whole obsession with metrics is entirely one-sided. An analysis function could detect the word "Bibliography" as a heading and apply a lower weighting to all the text beneath it. An analysis function could also parse the text, scan on-line dictionaries or a web search engine, reach a primitive conclusion on the incidence of obscure terminology - and apply a "difficulty" weighting to the text. Now, why do I have the feeling that the former function is more likely to appear in an analysis tool than the latter?

Considering that, I'm inclined against a standard because of its potential for abuse and for propagating misguided views about translation. If a standard is to be used, though, I consider it essential that:

- it be vendor-neutral. A "standard" would say that, for instance, a word is delimited by spaces (e.g. "do not" is two words, "don't" is one). Or that numbers in a table cell containing no text are not to be counted. Or that a standard line has 55 keystrokes. It should not say "the count shall be that yielded by analysis tool xyz". Firstly, the count algorithm of tool xyz is often not properly documented; secondly, it is a barrier to free trade (requiring purchase of a particular product); thirdly, it has no legitimate basis (see below). Software vendors would of course be free to design their analysis tools against a vendor-neutral standard. This doesn't just affect translation, incidentally; there's no consensus between word processing vendors over word count algorithms, and this has already led to court cases.

- its application should be voluntary. Translators should be free to choose to apply it, or not. I am in two minds about a standard, but personally I am strongly opposed to it being in any way compulsory.

- it should be drawn up by a recognized body. "The community" is not a recognized body, it is merely a group which is vulnerable to the influence of more powerful bodies, such as - in this case - translators' associations and software vendors. These bodies may or may not represent the stakeholders in the standard. If you take my view, that the charging method is the prerogative of the translator (which brings with it a responsibility on the part of the translator for it to be defensible), then translators' associations could arguably draw up such a standard; but it would be better still if a recognized standards body were to do so, which by definition means in consultation with the stakeholders.

Marc


Direct link Reply with quote
 
Pablo Roufogalis
Colombia
Local time: 07:38
English to Spanish
Trados is a must Feb 18, 2005

All productivity tools are a double-edge sword.

The moments when I think I can't do without Trados is when I don't remember a specially difficult word or phrase that I did before but I remember I have translated before. I used to agonize and invest a lot of time in situations like that like that but not anymore.

Also, it makes much more easier to create dual-language documents.

And there's nothing to discuss, those tools are here to stay.


Direct link Reply with quote
 
Pages in topic:   [1 2] >


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

Are we cheating ourselves by using analysis tools?

Advanced search


Translation news





LSP.expert
You’re a freelance translator? LSP.expert helps you manage your daily translation jobs. It’s easy, fast and secure.

How about you start tracking translation jobs and sending invoices in minutes? You can also manage your clients and generate reports about your business activities. So you always keep a clear view on your planning, AND you get a free 30 day trial period!

More info »
CafeTran Espresso
You've never met a CAT tool this clever!

Translate faster & easier, using a sophisticated CAT tool built by a translator / developer. Accept jobs from clients who use SDL Trados, MemoQ, Wordfast & major CAT tools. Download and start using CafeTran Espresso -- for free

More info »



All of ProZ.com
  • All of ProZ.com
  • Term search
  • Jobs