| Out of the rat race || May 5, 2017 |
Mario Chavez wrote:
Sight unseen, I cannot say yea or nay about Slate (I just saw a translator write a posting about assembling his own Slate setup with a CAT tool).
I wonder if we (translators, software vendors, MT and CAT developers) haven't become more obsessed with speed than with good writing. Quality models as they are (ISO or non-ISO) can't be applied or forced on writing. Writing is not coding or decoding.
I stopped translating on speed pressures years ago. I prefer to translate fewer words per day than sacrifice the precious time I need to read the original and write a suitable sentence or paragraph. Sure, I don't make as much money as before or as other guys claiming to make up to 10,000 words a day. I am absolutely not interested in becoming a cog and forgetting to enjoy what I do.
Doing translations for speed makes me think of this other model: translating for money.
Hi Mario, I hope it's okay to jump in here as owner of Slate Desktop and address your two points, translator postings and speed.
Re speed (reverse order)
I'm only aware of Loek van Kooten's review that talks about speed. This was important to him but none of our customers have raised this as a motivating factor in their use of SD.
We've done several A/B tests about how to describe SD. I don't recall any of them claiming speed improvements, but if one did it was a failed test. SD is definitely not about speed. Rather it's an efficiency tool... dare I say "productivity" but I think that's a tired, over-used term. One catch-phrase we have used is "finish more work in less time." This might imply working quickly, but that's not what we mean.
In our content, you'll see many claims as to SD's potential to create "correct" drafts that you accept as final without changing them, i.e. edit distance zero (ED0) segments. Every ED0 segment means you exercised cognitive energy to read, comprehend and judge a segment. It's closer to proofreading, i.e. no editing. That means you didn't do anything mechanical like typing or speaking (into Dragon) to edit. For most people, editing mechanics are slower than thinking Therefore, you don't actually type or speak more quickly to "finished in less time." You finish earlier because you reduce the mechanics.
Now, that's the ideal. Like many ideals, real life is (most often) different. In practical terms, each translator finds his/her own personal balance regarding actual habits that work best. The work habits that Igor (below) adopts might not be the same as those that Emma (below) uses. Therefore, it's very possible that you could use SD within your fewer words per day work habit and still finish more work in less time.
This article describes our recommended evaluation procedures to help you quickly (efficiently not speedily) test SD to determine if it's something you can learn to use and realize a benefit. Evaluating Slate Desktop By starting with these procedures, you'll probably develop your own stride in using SD, or you'll ask for a refund.
Re translator postings
To be honest, more than a year after releasing SD, I'm disappointed that more customers haven't publicly talked about their experiences. I don't know why and any reasons I might offer would be speculation. I can tell you that private tech support exchanges with customers are positive. They report they're benefiting from SD and they continue using it.
Here's the catch. Customer privacy is our prime directive. SD is fundamentally about respecting customers' confidentiality. So, I don't beg customers to publicly disclose their experiences. It's not their responsibility to do my advertising. I don't violate their trust. We're grateful for those who have provided testimonials. We don't even hold back reports that SD wasn't beneficial. It's important that new customers know that we'll fully support them during their trials and we will respect their decisions no matter what.
I've only found two critical reviews of people reporting SD was not beneficial. If you know of others, I ask that you share.
- Loek van Kooten tested SD for gaming localization. After I worked with him for a couple of weeks, he concluded that his highly customized memoQ segment assembly dictionaries worked better for him. Automated tests showed them neck-and-neck, but his personal experience was what mattered. We respected his choice and he received a refund within hours of his final decision. His complete experience is linked from this link.
- Michael Beijer has been quite vocal here on this forum discussion. He was one of our first supporters during our Indiegogo crowdfunding campaign when we pre-sold SD with a no-refund policy. Michael has been very honest and professional about his disappointment that SD with his TMs did not outperform Google Translate for his genre. We didn't refund his contribution, but I worked with him personally to help him recover his costs. Thank you Michael for your efforts.
These are three of our most visible customers. You can find others on our website.
- Igor Goldfarb's testimonial strongly credits SD with helping him meet tight turnaround times. In my personal exchanges with him, I get the sense that he's talking about efficiency not hastened speed.
- Pieter Beens has written several blogs about his experiences. They're linked from this link. I remember one of his other blogs polled translators about their per-word rates, but his post about SD doesn't connect the two. Maybe it's best to ask him if they're related.
- Emma Goldsmith has also written some blogs, also linked from this one. Again, speed is not been a major advantage for her.
I hope this helps and thanks for raising the issue of speedy work.
[Edited at 2017-05-05 12:58 GMT]
| || |