A lively honyaku

[Dan Lucas]

The recently locked thread on gender and so on was the liveliest we have seen for some time, whatever the regrettable vehemence of some of the opinions expressed. It proves that people do read honyaku, and can respond when something catches their attention. And yet the number of posts is usually very low.

Is this just demographics? A number of people have mentioned retirement or transition to other areas of work, and we've had one or two people pass away over the past few years

Are we seeing the bulge of Japan hands who got interested in the language back in the 1960s and 1970s leaving the stage, without anybody replacing them?

Or are they being replaced by younger people with no memory of the pioneering pre-internet days of Bix, the Well, Cix in the UK, or usenet, and for whom mailing lists and email in general are not an amazing convenience but rather a slow and fiddly annoyance compared to the immediacy of WhatsApp and similar apps?

Regards,

Dan Lucas

[Herman]

On 9/6/23 10:40, Dan Lucas wrote:
> The recently locked thread on gender and so on was the liveliest we have
> seen for some time, whatever the regrettable vehemence of some of the
> opinions expressed. It proves that people do read honyaku, and can
> respond when something catches their attention. And yet the number of
> posts is usually very low.
>
> Is this just demographics? A number of people have mentioned retirement
> or transition to other areas of work, and we've had one or two people
> pass away over the past few years
>
> Are we seeing the bulge of Japan hands who got interested in the
> language back in the 1960s and 1970s leaving the stage, without anybody
> replacing them?
>

I think one factor may be AI. Although the term itself is an
abbreviation of "Artificial Intelligence", AI signifies a
non-intelligent system that performs functions traditionally considered
to require intelligence.

In the field of translation, there has been much talk about whether AI
can reach the level of human translation, but what seemed significant to
me, in terms of the impact of AI on translation, apart from the
commercial aspects and impacts on human translators, is that, in my
understanding anyhow, what would happen is not that AI would approach or
achieve some standard of human translation, but rather that the realm of
translation as a whole would shift toward the AI standard (while
probably maintaining a human niche).

So far, it seems that is exactly what has been happening.

For example, there are at least some mid-level information providers who
regularly include MT-generated text as part of their content (such as
news-related content), where the translation contains not just errors or
inaccuracies vis-à-vis the source language, but also a notable amount of
text that would have previously been perceived as nonsensical. However,
if one regularly consumes text content that includes, e.g. terms that
make no sense in the given context, it would be expected, as an
adaptation to the exposure to such content, that one would cease
perceiving these as errors or irregularities but rather as expected
features of the text, and concurrently, that one would not take the
individual meanings of individual words to be necessarily decisive or
significant to understanding a text as a whole.

Thus, the question of what is the exact meaning of some text or term, or
what is the correct or most accurate way to translate something, ceases
to have the relevance that it once did.

And insofar as this mailing list was dedicated to asking and answering
just such questions, it can be expected that the number of posts would
decrease as the relevance of such questions to translation decreases.

Herman Kahn

[Jan Cash]

As one of the "newer" translators inching toward a decade of experience in the field, I think the issue is a problem localized to this mailing list. The non-translator lists I've joined are active on a daily basis, and I also know of several specialized translator groups (though not mailing lists) that grow each year. I can't speak for others, but I've considered leaving this group in the past because the posts often come across as unprofessional and political.

Dan, whether it's intended or not, you've even phrased this current topic in a way that borders on insulting to new professionals, which was unnecessary for the question you were asking. I'm not sure if the handful of members who speak to others this way truly don't realize how they come across, but I also don't believe anyone will reflect on their behavior or attempt to be more personable if asked.

Regardless, whenever I read a topic, I more often than not find the posts aren't relevant to actual work as a translator, or are phrased to get a rise out of other members. I believe that's the reason why "younger people" don't want to engage or stick around.

Sincerely,

Jan Mitsuko Cash

[Joe Jones]

I've been translating for about 20 years (on the side of practicing law for the last 15) and sometimes get the impression that I may still be on the junior end of the demographics here.

I basically never ask language questions here because I can get quicker and more reliable answers from a Google or Linguee search. But I find the discussions to be very interesting, especially the perspectives of the older folks. Some of you wrote the books that I used to learn Japanese in the late 90s, and I am always happy to know that you are still around and contributing to everyone's understanding. The recent discussions about AI and the future of the profession have especially piqued a lot of thought on my part.

[Sandra Ogata]

On 9/6/23 10:40, Dan Lucas wrote:

> The recently locked thread on gender and so on was the liveliest we have
> seen for some time, whatever the regrettable vehemence of some of the
> opinions expressed. It proves that people do read honyaku, and can
> respond when something catches their attention. And yet the number of
> posts is usually very low.
>
> Is this just demographics? A number of people have mentioned retirement
> or transition to other areas of work, and we've had one or two people
> pass away over the past few years
>
> Are we seeing the bulge of Japan hands who got interested in the
> language back in the 1960s and 1970s leaving the stage, without anybody
> replacing them?

>

I think there are many members like myself, who read the posts with interest and hope to be of assistance at some point.

The GFC was, for me, a critical changing point in my translation. Translation dropped off and I found myself having to look for other income sources, which took me back into teaching.

I still translate, though in a very limited sense, with a very limited subject matter compared to before the GST (and child raising). the pandemic and its effect on the airline industry has further limited the level of incoming work. About 6 months ago, I was told the contract would end, but happily, it has continued, albeit in a far diminished form. Every month though, I expect the email saying they are resorting to AI....

While I stay in the industry, though, I will continue to lurk and hope that I can contribute and answer to one of the questions posted.

I thank the core of the members who regularly post!

Sandra Ogata

Brisbane, Australia

--
You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/3e3fd56a-719f-47e0-b1ec-ea2447ac2d8b%40app.fastmail.com.

--

SOS Translations--For all your translation needs
Sandra Ogata
46 Ellington Street, Tarragindi, QLD, 4121
Australia
Tel: 61-7-3848-8595
Mob: 61-410-038-572
Email: shan...@gmail.com

[Brendan Craine]

I probably can be said to fall into the demographic of "young people," and I certainly would be considered a less-experienced translator. I didn't necessarily find Dan's phrasing insulting, as Jan suggested, but I would agree that this listserv does lack a general sense of warmth and amity. I prefer to lurk rather than participate because I have the sense (rightly or wrongly) that any assertions I make, advice I offer, or opinions I share are being scrutinized for flaws. There is a certain amount of one-upsmanship that goes on in the threads. 

Or, in many other cases, I simply don't have the technical knowledge to be of any use to, for example, the patent translators of the group.

I am also part of a translation-focused Discord server, and it is much more active. Yes, the overall age demographic does seem to skew lower, but there's also a lot more good-natured bavardage and open discussion, such that I feel like my inexperienced voice still carries some value. There is also more sharing of resources, industry gossip, job posting, and general conviviality than I see here. 

So, although it's true that email lacks some of the immediacy and functionality of a service like Discord, I don't think that's why I (or anyone else of my age group) are not more participatory. 

I would be curious to hear whether any other lurking 30-year-olds feel similarly. 

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/97d250d1-3a8d-4e27-8a91-2cafcf29e761n%40googlegroups.com.

[Bill Lise]

I think the recently locked thread seemed be so lively because the normal baseline level of activity in this group is very low.

Most of the online interaction between JA/EN translators long ago migrated to social media platforms. There is a JA/EN translators group on Facebook, for example, which has about 800 members and often has multiple threads starting on any given day. I think another factor that might be at work is that many of the translators I interacted with in that group are people I personally know in the real world and some who I meet with socially. I sense that there is a greater desire to interact with people you know in the real world. 

Not wanting to continue to contribute to Meta's business model, I left Facebook some time ago, but there appears to be nothing else as an active online venue for JA/EN translator interaction. I could be wrong. JAT formerly had a mailing list, but that was replaced by difficult-to-use forums that had domain specificity.

I am hoping that JAT can return to a single, easy-to-use online venue that is not linked to a social media platform.

On the issue of aging, being one of the oldest of the old farts, I would just say that I don't necessarily think that the lack of activity is because of a significant demise or retirement of elderly translators. Translators are lucky in that they can work well into their senior years, adjusting work volume as energy permits.

As a case in point, after I had passed 70, I received and executed my largest interpreting job ever--35-plus consecutive days in the Nissan Ghosn/Kelly case, at the behest of prosecutors office in late 2018; Google is your friend--and my largest single translation job ever (very recently, 1097 pages of Japanese police documents for a US Navy court martial.

Regarding attrition from our ranks, we will see some people leave because of age, but there is another wave of attrition that is happening and not getting much attention. That is the people leaving because they cannot make a living from--or cannot stomache having to do--MTPE and don't have work sources other than agencies that are rapidly adopting the MTPE business model. That attrition will accelerate.

Bill Lise

[Bill Lise]

The question about translator attrition, including older translators dying off prompted me to look back at a collection of comments from translators that Tom Gally created, drawing on comments from a large number of translators (including me) before and after the turn of the millenium.

https://www.gally.net/translation/gettingstarted.htm

I was surprised to see many translator names I knew but cannot recall hearing from or about for many years. Many were young enough at the time to still be translating now. Does this mean that some translators have become recluses? Or perhaps they have found other careers?

The shrinking of our ranks continues, but the current attrition is for reasons that are more easiliy understood.
Bill Lise

[Dan Lucas]

I won't comment on individual posts - we've just had one rather fiery thread and I don't want to inadvertently catalyse another - but there have been some interesting perspectives in the responses.

Thank you

Dan Lucas

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.

To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/3e3fd56a-719f-47e0-b1ec-ea2447ac2d8b%40app.fastmail.com.

[Karen Sandness]

Regards,

Karen Sandness

Here’s my take on it as someone who joined Honyaku in 1994.

While it’s true that some of the most prominent participants have died, retired, or otherwise left the profession, it seems that the action has moved over to Facebook.

I know that when I have a question that needs to be answered *right now,* I’m more likely to receive a timely answer in the Facebook Translators (Japanese <->English) Group, which includes participants from the Americas, Europe, Asia, and Australia or New Zealand, than I am from Honyaku. No matter when I post, *somebody* among the 744 NJS, NES, and N-something else-S is bound to be awake and able to answer.

I don’t know how many people are subscribed to Honyaku, because even the page under Google Groups doesn’t give that information, but there are very few initial posts, and they’re all from the same half-dozen people. 

For better or worse, Facebook is where the action is currently.

Historical perspectively yours,

Karen Sandness

On Sep 7, 2023, at 07:00, hon...@googlegroups.com wrote:

A lively honyaku

"Dan Lucas" <dan....@carninglipartners.com>: Sep 06 06:40PM +0100

The recently locked thread on gender and so on was the liveliest we have seen for some time, whatever the regrettable vehemence of some of the opinions expressed. It proves that people do read honyaku, and can respond when something catches their attention. And yet the number of posts is usually very low.   Is this just demographics? A number of people have mentioned retirement or transition to other areas of work, and we've had one or two people pass away over the past few years   Are we seeing the bulge of Japan hands who got interested in the language back in the 1960s and 1970s leaving the stage, without anybody replacing them?

Or are they being replaced by younger people with no memory of the pioneering pre-internet days of Bix <https://en.wikipedia.org/wiki/Byte_Information_Exchange>, the Well <https://en.wikipedia.org/wiki/The_WELL>, Cix <https://en.wikipedia.org/wiki/CIX_(website)> in the UK, or usenet, and for whom mailing lists and email in general are not an amazing convenience but rather a slow and fiddly annoyance compared to the immediacy of WhatsApp and similar apps?   Regards, Dan Lucas

[cpta...@ozemail.com.au]

I wouldn’t take “presence on Honyaku” as a particularly reliable indicator of translator population. I think that the relationship people have with this or that form of dopamine dispensing app goes through an arc or a life cycle. I was madly active on Honyaku back in the day, then you kind of run out of energy, then I did it with FB, and then left FB altogether ironically because of the censorship. It too was run by scolds who wanted some sterile, idealised, perhaps outdated vision of what “professional” discourse was. It wasn’t lively enough for me basically.

Anyone who thinks that in order to fulfil their professional duties translators’ discussion won’t inevitably drag them to and through passionate debate about every topic under the sun, well, I don’t think they understand translation frankly.

Chris

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/3804D122-A02B-491A-AFA9-F27924657929%40gmail.com.

[Jon Johanning]

Hi Karen,

The "Translators (Japanese <->English) Group" you mention is a new one to me, and searching for it on Facebook, I don't turn it up. But I'd like to join it. I'm out of the loop on lots of things right now, I realize.

On the subject of the recent Honyaku list traffic, I'm not posting a lot, because I haven't been getting a lot of problems that I can't solve by using DeepL (with all of its imperfections) and other on-line resources. On the recent trans thread, I didn't feel inspired to contribute anything (and I only skimmed it a bit because I didn't have time to read every post carefully). 

Though getting older by the minute (at the customary rate of 1 min/1 min), I'm not dead yet, and would like to continue using this group when I have questions to ask or answers to give. If the action is now elsewhere, and all the kool kids have flown this coup, I'm certainly interested in following the flock.

Jon Johanning

zenn...@icloud.com

[Bill Lise]

That Facebook group was made secret, which is why it does not show up. Because of militias, kidiporn, and more, I think Facebook has withdrawn that option. 

You can be invited to the group, and Karen can do that. I’m gone from Facebook, so I can’t. 

Bill Lise

--
You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/85e7e12c-c955-4221-bc6e-afc7ebbd595fn%40googlegroups.com.

[Jan Cash]

Anyone who thinks that in order to fulfil their professional duties translators’ discussion won’t inevitably drag them to and through passionate debate about every topic under the sun, well, I don’t think they understand translation frankly.

Chris's words, bolding mine.

I'm part of other groups that are able to discuss sensitive topics while being professional, without taking jabs at others in the process. I don't feel comfortable posting in this group (most of the time) because of these sorts of comments. It's not necessary for the conversation, the purpose seems to be to get a rise out of members who don't agree with the poster, and they're alienating.

While Brendan didn't agree with my earlier post, he did mention finding that the listserve lacked amity and I think this is a case in point.

I personally would like to talk to people from diverse backgrounds, but I usually have to seek out people outside my age group in platforms and places they're comfortable in (both older and younger generations). That's why I joined this list, but I've never felt welcome here. If people leave because they're antagonized whenever they post an opinion that's not agreed with, this listserve will just become a place for similarly minded people, and ironically, there won't be much true debate happening then.

Jon, Bill already mentioned the Facebook group, but there are a few other places that might be interesting to you too.

If you're interested in Discord, there's a general translation server that's open to everyone. 

Here's the Reddit thread on that: https://www.reddit.com/r/TranslationStudies/comments/hqt0np/discord_server_professional_translators/

If you specialize in localizing popular media, there's the Honyaks Discord, which can only be joined by application: https://www.hon-yaks.com/about-us

I run a Discord for editors working in localization of popular media (mostly manga, novels, games). Anyone's free to join, but I only share invitations on an individual basis.

There's also a literary translators' listserve here: https://emerginglittransnetworkamerica.wordpress.com/

It's not specific to Japanese<->English. I believe there's also a version of this group for European translators.

If you don't mind paying dues, there's also SWET's listserve, and I think JAT has a bulletin board.

Best,

Jan Mitsuko Cash 

[Bill Lise]

Jan

>>>>

If you don't mind paying dues, there's also SWET's listserve, and I think JAT has a bulletin board.<<<<

My understanding, confirmed by people in JAT, including current and past directors, is that JAT no longer has an actively used mailing list or other forum as they did before. Years ago they started subject-specific forums, which I’ve been told are difficult to use. 

And while all that was happening social media took over. One reason is addiction to a specific social media platform. Another is that the platforms are _ostensibly_ free. 

There appears to be a prevailing idea that everything should be free and the feee version often defeats a paid-for version. 

Although I was a founding member and multiple-term director/president of JAT, I left the organization years ago.

However, if they rebuilt a JAT-run online venue that is easy to use, I would rejoin in a flash. The dues (10,000 yen?) are small change. 

Bill Lise

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CADXUeAxK2M_LEG6jWc0N3015zqN5fUPkWkq9nnZZyaG_t8-cGw%40mail.gmail.com.

[Jan Cash]

Bill, you're definitely right about JAT's current system. I have email notifications set up for the forum/board/whatchamacallit they do currently have, but I've never posted and I usually don't read the posts because of the format. 

SWET, which is a little broader in scope (the WE stands for writers/editors), does have a mailing list and at least one specialized group Discord. They mostly post about events, but I find that I do read most of the emails, even when the topics don't seem like they'd be relevant to my work.

[Oroszlany Balazs]

Jan, thank you for these suggestions!

It looks like there are plenty of communities, but no central hub - and, for a while, this mailing list was similar to that. 

If you need a Reddit account in order to get a Discord invitation, or (for perfectly valid reasons) the most lively discussions are in hidden FB groups, then you need luck and connections to find these.

(I have realised how difficult it is to rebuild this kind of network when, a few years ago, I moved to the UK. Slowly changing your clients is one thing, but finding local colleagues, not necessarily in the same language pair, who are dealing with similar challenges, is much harder. Maybe they are there - but on a Mastodon server, a Discord channel, or on a social media site you haven't heard about.)

Balazs

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CADXUeAxK2M_LEG6jWc0N3015zqN5fUPkWkq9nnZZyaG_t8-cGw%40mail.gmail.com.

[cpta...@ozemail.com.au]

“..the purpose seems to be to get a rise out of members who don't agree with the poster, and they're alienating.”

 

You’re right. I apologise. It worked though. But now I feel bad.

It can also be the case that I genuinely want to have the discussions too. And as I consider translators to be gods rather than slaves of language, I guess I expect them to be able to handle any topic in any register without “taking offense”. But maybe I am being unrealistic.

Chris

 

 

From: hon...@googlegroups.com <hon...@googlegroups.com> On Behalf Of Jan Cash
Sent: Friday, September 8, 2023 8:40 AM
To: hon...@googlegroups.com
Subject: Re: A lively honyaku

 

Anyone who thinks that in order to fulfil their professional duties translators’ discussion won’t inevitably drag them to and through passionate debate about every topic under the sun, well, I don’t think they understand translation frankly.

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CADXUeAxK2M_LEG6jWc0N3015zqN5fUPkWkq9nnZZyaG_t8-cGw%40mail.gmail.com.

[Bill Lise]

Depending upon individual circumstances and geography (yes, geography, even with the Internet), it can be very difficult to establish a network of real-world colleagues or clients outside of cyberspace.

JA-EN translators living in major urban areas of Japan are fortunate in having the translation population critical mass density to enable face-to-face meetings, and many of the people I interact with offline are people I met when JAT had monthly FTF meetings, which were cut back way before the pandemic struck.

If I were to move to the US (highly unlikely, of course, after being here for nearly a half-century), I would be at a loss to connect in the real world with colleagues and, more seriously, if I didn't already have numerous direct clients there, I wouldn't have a clue as to how to develop direct clients.

The clients I have in the US are almost all law firms, and I have them as translation clients because I met them face-to-face while interpreting in depositions here (sometimes on the "enemy" side).

With agencies moving to MTPE, network building for sales is increasing in importance. I don't have a universally applicable solution. In fact, if I didn't already have a network, I wouldn't even have a solution for myself.

Bill Lise

[John Stroman]

I'm one of the older members since I first joined this list in 1989 when I became a freelance J>E translator. At that time I was overjoyed when I discovered there is actually a list of professional J<->E translators where I could ask language or social questions and discuss the business of translating. In the US at that time, the Apple Macintosh could display Japanese with a word processing package called EGワード, but it could not connect via a modem. My PC could use a modem, but could not display Japanese. We translated from paper documents that arrived either by post or fax printed on thermosensitive paper. Most of the documents had been photocopied first and were in 4 point font.

When faced with a bunch of illegible marks on a page, we could head for our bookcases full of weight-lifting tomes such as the Green Goddess, Brown Behemoth, Nelson, and the IPC technical dictionaries. Honyaku saved the day for me on many occasions, and although some of us could not display those blurry kanji on a PC, we could still figure out their identity through 何々冠、何々偏 and so on together with some context. In many ways, whether living in Japan or elsewhere, we were lucky to have a group of people who were willing to help in deciphering an exotic language that was in very high demand, or in return, provide background information for Japanese documents about the US when their content was confusing to translators living in Japan. Younger translators probably cannot envision the conditions we worked under, but they were as modern as technology allowed at that time.

Because freelance translating is a solitary occupation, the Honyaku mailing list also served as our social media platform, and vicariously we became friendly and personal with many people we could never expect to meet face-to-face during our lifetime. That may explain why some of the back-and-forth on the list appears to younger translators as unprofessional. Posters who express matters of personal opinion write under the assumption that we all belong to the same community working toward common goals. Although members will sometimes disagree, we belong to the same group and will help each other out if asked despite differences in personality and temperament.

Younger translators are always welcome.

John Stroman

----------------

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/87fedc16-3e78-44a7-8b8d-c3acfd42a92dn%40googlegroups.com.

[cpta...@ozemail.com.au]

One of the reasons FB took off is because it has a “like” button, for posts like John’s..

Chris

 

From: hon...@googlegroups.com <hon...@googlegroups.com> On Behalf Of John Stroman
Sent: Saturday, September 9, 2023 3:56 AM
To: hon...@googlegroups.com
Subject: Re: A lively honyaku

 

I'm one of the older members since I first joined this list in 1989 when I became a freelance J>E translator. At that time I was overjoyed when I discovered there is actually a list of professional J<->E translators where I could ask language or social questions and discuss the business of translating. In the US at that time, the Apple Macintosh could display Japanese with a word processing package called EGワード, but it could not connect via a modem. My PC could use a modem, but could not display Japanese. We translated from paper documents that arrived either by post or fax printed on thermosensitive paper. Most of the documents had been photocopied first and were in 4 point font.

----------------

 

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CAJVTx95g%2B64XVKfw3kT9WB%2BwXwNgmiH-0WvBqLaqgCHHDXOtig%40mail.gmail.com.

[timl...@aol.com]

Thankyou John, very much my experience too.

Tim Leeney

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CAJVTx95g%2B64XVKfw3kT9WB%2BwXwNgmiH-0WvBqLaqgCHHDXOtig%40mail.gmail.com.

[JimBreen]

Karen Sandness wrote:

> While it’s true that some of the most prominent participants have died, retired, or otherwise left the profession, > it seems that the action has moved over to Facebook.

This is true. I think it's a pity but it's what has happened in many forums.

As an observer of Japanese translation rather than an active player, I found the Honyaku list, which I've read via the daily summary of postings, to be a handy on-stop-shop in my inbox. The problem with the Facebook group is that I have to keep it open in a tab and monitor it, including digging into follow-up comments.  Sometimes it's useful - I'm about to add ラーケーション to JMdict following a recent FB post by Joji Matsuo. I'd still prefer them landing in my inbox.

Jim

[Bill Lise]

I think what removed content (members) from this group and effectively killed the online venue(s) of JAT was Facebook.

People were made to believe that Facebook is aboslutely essential and (particularly disadvantageous for JAT), Facebook is thought to be free.

Some people might learn that Facebook is not essential (I have learned that---83 days free and counting), but the fact that there is no _cash_ changing hands to use Facebook places JAT at a great disadvantage in a world in which many people are loathe to pay for services. The Internet made everything free, right? Wrong, but that is a topic for another time.

[JamesALockhart]

> While it’s true that some of the most prominent participants have died, retired, or otherwise left the profession, it seems that the action has moved over to Facebook.

Yup.

Seconding thatly yours,

—Jim Lockhart @it since 1984

[JamesALockhart]

Great times, though; weren't they, John? :)

And for the information of those who are averse to heated exchanges: I made some pretty good friends and found some pretty good colleagues through those arguments. Steve Venti, Wolfgang Bechstein (rest his soul), Karen Sandness, and Bill Lise are four names that come to mind. They are all guys I argued rather vehemently with at one time or another, and they are also people who have been indispensable to my career and personal development. So rather than eschewing it, I recommend that you "young people" learn how to do it constructively, without taking disagreement personally.

Reminiscingly yours,

—Jim Lockhart

[Joe Jones]

On Saturday, September 9, 2023 at 9:59:47 PM UTC+9 JamesALockhart wrote:

And for the information of those who are averse to heated exchanges: I made some pretty good friends and found some pretty good colleagues through those arguments. Steve Venti, Wolfgang Bechstein (rest his soul), Karen Sandness, and Bill Lise are four names that come to mind. They are all guys I argued rather vehemently with at one time or another, and they are also people who have been indispensable to my career and personal development. So rather than eschewing it, I recommend that you "young people" learn how to do it constructively, without taking disagreement personally.

And I would recommend that everyone who makes generalizations about large, diverse groups of people... just stop.

There has been too much of that in this group in the last couple of weeks.

[Jan Cash]

I'm not sure this list is very diverse though. There was even a comment in another thread that the list is probably mostly white and old. There's also a difference between allowing anyone to join and welcoming others to be here. 

For example, are there trans people in this group after the gender vs sex thread? If there aren't, there's also a possibility others have been silently self-selecting out. I remember interesting topics coming up and more people participating in discussions a few years ago--I'm not sure whether they're still here. Maybe they are and they just don't post.

I also don't understand the appeal of being unpleasant to strangers on the internet or giving someone free license to do so by calling it a difference in temperament or personality. 

I've felt for the last few months that this place isn't for me and I don't know that there's much value to me sticking around, but I wanted to bring up some things I've heard a lot elsewhere and haven't seen as much in conversations here. Or maybe these have come up and I just missed the discussions. In any case, I felt that some of these things would be sensitive topics for people who had different experiences and I didn't want to add to anyone's distress while they were struggling.

-Complaints about being overworked/having too much work. Every few months, I refer one to two new translators to clients because they need more, so there's still high demand in some fields.

-Quite a few of my clients or their employees are trans or nonbinary. Some of the opinions expressed in other posts probably would make a person unemployable since they're a harassment risk.

-There's some concern about AI in my field, but for the most part agencies and companies with lower rates have been the only ones attempting to integrate MTL into their workflows. The "high-quality" clients haven't made any moves to use AI, and I think the people employed at my clients would revolt against it since they know it would add to their workloads. 

-Very little MTL editing. I've been asked to evaluate MTL products and I've heard gossip about people being offered some MTLE work, though I think that was in other fields, not mine. I haven't been offered MTLE work at all.

-Client complaints about a lack of novel translators in particular. For transparency, rates for these novels are very low. A 300-page volume would be about $3000-$4000, but they're also steady income. I've never experienced a famine period as a full-time freelancer since I started working on them.

Jan Mitsuko Cash

[cpta...@ozemail.com.au]

“There was even a comment in another thread that the list is probably mostly white and old.”

 

Yes but thanks to colonialism and imperialism old white men DO constitute a repository of all that diverse cultures have to offer, so you know, all good.

Chris

 

From: hon...@googlegroups.com <hon...@googlegroups.com> On Behalf Of Jan Cash
Sent: Sunday, September 10, 2023 4:58 AM
To: hon...@googlegroups.com
Subject: Re: A lively honyaku

 

I'm not sure this list is very diverse though. There was even a comment in another thread that the list is probably mostly white and old. There's also a difference between allowing anyone to join and welcoming others to be here. 

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CADXUeAzgA6angBpRCzBnnFsrwAgk221zD%2BawsWd%2Bo4-J_OEwFg%40mail.gmail.com.

[Jan Cash]

Chris, I didn't think I'd get a reaction like this from you with just a bit of prodding or that you'd be so direct. It's given me peace of mind that I judged your character correctly, so I suppose that's a silver lining.

When my grandmother's memory started to go and she switched from speaking standard Japanese to mostly Uchinaguchi, I had a lot of trouble trying to communicate with her. That language barrier was a direct result of the occupation, so no, imperialism and colonialism don't make groups diverse.

In my opinion, what you said sounds racist. Even if you are not a bigot and simply write these things for the reaction or just to be controversial, I don't want to be associated with a group that accepts someone like you, so this will be my last message and I will no longer recommend this listserve to other people.

Sincerely,

Jan Mitsuko Cash

[Dan Lucas]

On Sat, 9 Sep 2023, at 19:57, Jan Cash wrote:

I'm not sure this list is very diverse though. There was even a comment in another thread that the list is probably mostly white and old.

FWIW I agree with you. The honyaku list does seem to me to skew older and whiter, probably quite significantly. Given that social media in general (Facebook aside) skews heavily younger and more progressive, this list is probably unusual. In that sense it arguably adds considerably to the diversity of JA<->EN discussions online.

I also don't understand the appeal of being unpleasant to strangers on the internet or giving someone free license to do so by calling it a difference in temperament or personality. 

You see, this would be an ideal opportunity for me to mirror your comment earlier in the thread about "borders on insulting" by making the point that you are borderline accusing other commenters of being unpleasant.

I could, but I won't, because I don't let myself get worked up about such things. Speaking in general terms, if one goes looking for reasons to be offended one will always find something because such judgements are innately subjective.

For my part, I do not go out there actively seeking to offend people - I try to use moderate language - but neither do I police myself down to the nth degree.

The normal cut and thrust of discussion is going to involve clashes of opinions and a variety of modes of expression. As was pointed out in a recent case here in the UK, mocking or satirising the opposing view is part of the common currency of debate - and probably always has been.

Compare that to the common exhortation of "be kind". In the previous thread, for example, it became clear that no compromise was acceptable to some participants: "being kind" meant agreeing with their point of view, period. There was some heated language on both sides.

In such cases, what does one do? One school of thought is that we simply avoid any difficult conversations. I disagree. I argue that we should open up the shutters and let in the sunlight. Eventually the shadows will be driven away.

To take a parallel example, I do not believe that we can eliminate racism - an ongoing concern for me given that my children are of mixed ethnicity - by closing down all discussion of racism that does not meet some moderator's preconceptions of acceptability.

I've felt for the last few months that this place isn't for me and I don't know that there's much value to me sticking around,

Entirely your choice, of course. You have agency.

but I wanted to bring up some things I've heard a lot elsewhere and haven't seen as much in conversations here. Or maybe these have come up and I just missed the discussions. In any case, I felt that some of these things would be sensitive topics for people who had different experiences and I didn't want to add to anyone's distress while they were struggling.

You talked about self-selection earlier. Here is its disreputable cousin, self-censorship. You're not doing the world a favour by refraining from discussing such matters.

Silence will not resolve the problems of those who have insufficient work and those who might have benefited from your comments will not have done so. You have a different perspective that some may find valuable. By not expressing that perspective you have further narrowed the scope of discourse. Net-net, that's a negative.

If you were metaphorically prancing around waving bunches of crisp $100 bills in the face of those who are struggling while singing "Eat your hearts out, poverty-stricken old white guys" that would be considered somewhat mean by some, sure.

As it happens (a) I don't think you would do that and (b) I certainly wouldn't seek to censor your speech even if you did, though if I disagreed with you I would deploy counter-arguments as I saw fit.

-Complaints about being overworked/having too much work. Every few months, I refer one to two new translators to clients because they need more, so there's still high demand in some fields.

There is indeed. I have a similar "problem". However, with two teenage sons to feed, clothe and educate I grit my teeth and accept larger projects with challenging deadlines more often than I would like.

On the flip side, it makes me appreciate downtime in the garden or woodland even more. And at my age a freelancer's life is still preferable to working in a large multinational. Overall, I have no regrets.

-Quite a few of my clients or their employees are trans or nonbinary.

That's interesting. I don't know if any of my clients are trans or non-binary.

In fact, in many cases I don't even know their sex.

It only recently dawned on me that one project manager with whom I often communicate by email is probably male, though I have thought of them as female for the past several years. I have no idea how any of them identify, but given that I use the same language for everybody, it's not an issue.

Some of the opinions expressed in other posts probably would make a person unemployable since they're a harassment risk.

I'm slightly worried that the moderators may interpret your comment as an attempt to reopen the previous thread by stealth. Still, at the risk of us jointly infuriating the mods I would point out that in the UK, where I live, this dynamic would cut both ways. Provocative statements by either side would be considered a harassment risk and detrimental to one's opportunities for employment (cf. the recent Sibyl Ruth case).

-There's some concern about AI in my field, but for the most part agencies and companies with lower rates have been the only ones attempting to integrate MTL into their workflows. The "high-quality" clients haven't made any moves to use AI, and I think the people employed at my clients would revolt against it since they know it would add to their workloads.

-Very little MTL editing. I've been asked to evaluate MTL products and I've heard gossip about people being offered some MTLE work, though I think that was in other fields, not mine. I haven't been offered MTLE work at all.

Agreed. I'm not seeing much if any MT or MTPE either, though I know that some clients continue to explore and experiment.

Nevertheless, CAT tools have cut out a lot of the boilerplate in my areas already. I suspect that finance and the grittier side of tech (usually industrial rather than software in my case) is far more susceptible to MT than, say, manga/anime or novels.

For the former in particular I think you need a good grasp of (current) demotic Japanese. That's hard to acquire and maintain, so a barrier to entry.

-Client complaints about a lack of novel translators in particular. For transparency, rates for these novels are very low. A 300-page volume would be about $3000-$4000, but they're also steady income. I've never experienced a famine period as a full-time freelancer since I started working on them.

I think it has always been the case that rates for literary translation, except for the authors with the very highest profiles, have been low. If we assume a book with 600 characters a page and 300 pages, that's 180k characters. At current exchange rates $4,000 is roughly 600,000 yen, so that's about 3 yen a character, which is frankly dire.

To put it another way, if one translates 500 characters an hour, it would take 360 hours to complete such a project, or roughly $11 an hour, vs an average of $13 an hour for Mcdonalds. Maybe that's the real reason that clients struggle to find translators. If clients were to raise their rates they'd get more interest, but maybe despite their complaints they don't need to?

Personally I have never been interested in literary translation, though I admire those who can do it well. Even if I were to attempt it (and assuming that I were to succeed, which I think would be highly unlikely given the breadth and depth of the competition) I don't think I could sustain my current standard of living.

But if you love the art and process of translation for its own sake, surely this is the ultimate niche in which to specialise. You forgo some unspecified economic upside because you simply love going to work every morning.

In contrast, I'm going to stop now because I need to complete by this evening 6,000 characters of corporate presentation material aimed at overseas investors. The thought doesn't set my soul ablaze with joy, but it's my own choice.

Dan Lucas

[Dan Lucas]

Although Jan will not see this, I apologise for not noticing her earlier email to Chris in which she stated that she was leaving.

I tend to hit reply to all, which of course included her personal email address.

I suggest that other posters, if any, are careful NOT to do what I just did.

Thanks

Dan Lucas

On Sun, 10 Sep 2023, at 09:35, Jan Cash wrote:

I already left this group. Please don't CC me in replies or contact me anymore. 

Thanks!

Jan

[timl...@aol.com]

Thanks Dan, much good sense.

Tim Leeney

--
You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.
To view this discussion on the web visit

https://groups.google.com/d/msgid/honyaku/1c627701-1e09-4f8d-959c-323c6020360a%40app.fastmail.com

.

[Carl Sullivan]

All,

Some good comments here. Interesting thread. Guess I will shoot one out.

I enjoy the topics from time to time, but am very seldom on here either because I am too damn busy translating, or I am just too off the wall for the mainstream. I do have a lot of friends here that I care greatly about--wish you well. 

I am as busy as could be still in the business. I see comments about the business drying up; I am sorry for that. It does not yet apply here. Japanese has slowed down a bit; Korean is insanely busy. I work MT postediting, etc. to contribute to my own eventual demise. Patent work is still big; other fields as well. I am translating for about 15 agencies, 7-8 constantly. 

I am a white, old, male, retired Navy WASP Irish-American who likes his beer, whiskey and powder skiing, living here behind Utah's Zion curtain in a cabin in the woods literally, doing this stuff. Maybe I qualify for some person of color lovin' for being married to a Japanese lady, with a Brady Bunch of Japanese-American kids and grandkids underneath us.

I don't think that folks are on here to offend or cause trouble. People nowadays are overly sensitive--get over it. Be grateful to God for each day you are given. I am truly grateful for what this field has given us. 

I wish all the best. 

Carl Sullivan

Former Administrator (when T-Rex was still roaming the earth), Japanese Division ATA

Sterling Utah USA 

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/367571050.7475816.1694335366403%40mail.yahoo.com.

--

Carl T. and Masae Yonaiyama Sullivan

MasaCa Translation Services LLC

masacatransl...@gmail.com

MasaCaTranslations.squarespace.com

801-369-1501

CDR Carl Thomas Sullivan Sr., US Navy Retired

Disabled Combat Veteran

[Bill Lise]

White and old here. 

About the market, I am busier than ever and have never once been asked to do MTPE.

The secret for me escaping the fate of MTPE is having only direct clients, although someday they also might start using MT, at which point even translation brokers will be in trouble.

That said, I know several translators who have left translation because they cannot make a living from MTPE, and I am pretty sure it will be getting much worse quite soon. Perhaps three years or so remains before most agency-dependent translators stop getting _translation_ work from agencies. 

I also see that numerous translators are openly offering MTPE on their websites and (less surprisingly) in their profiles on reverse-auction platforms like Proz, so I think the agency MTPE business model will succeed in securing enough post-editing labor. 

There are survival strategies for translators that avoid having to leave translation, but they’re not for everybody and not achievable by everybody. 

And regarding age, you can continue as an old white (or any other color) translator until quite an advanced age. 

As I _might_ have mentioned in another note, both my largest ever single interpreting assignment/invoice (to 特捜部 for interrogations of an executive of a certain auto company) and my largest single translation invoice (1097 pages of police documents for a USN court martial) happened well after passing 70. Age didn’t matter. 

Retirement is not in site. 

Bill Lise

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CA%2BQjO-kUXpGGdP_AwfqsWPFvsfurcOsFpB547WfNf6v_GVr73A%40mail.gmail.com.

[Bill Lise]

While retirement is not in sight, I do look forward to taking on just a bit less work. That said, it is somewhat difficult with direct clients, and it is a difficult task to hand off a Japanese direct client to someone else, since the client would normally expect to be able to interact with my replacement as they do with me. 

This reminds me of the several unintentionally comical sales emails I have received recently from two places that want to broker the sale of my company. They totally misunderstand what I do, and I would never aim to form a company that could be purchased successfully by someone else, like someone purchasing a customer list or the downline of a pyramid scheme. That’s not the kind of business I’m in. 

Bill Lise 

[Jon Johanning]

On several subjects recently mentioned:

As a mostly old, mostly white (or actually all both), non-trans guy, I've been happy with the list for a long time and still am.

I appreciated John Stroman's reminiscences about the old days very much. I got hooked on translating when the only way to do it (besides scratching on clay tablets or setting quill pen to parchment) was typewriting and delivering finished sheets of paper on foot or by snail mail. A little later, I often got faxes which were practically illegible (and not seldom in Chinese or Korean -- people couldn't tell one peculiar script from another). But I miss those days, because it was all so slow and leisurely a way to work.

Have machine "translation" and LLMs reduced work opportunities and led to list members quitting the profession? Certainly that is one factor in reducing the roster of list members. But that's been going on ever since Google Translate etc. came out with instant, free, and truly horrible "translations." Some clients have gone that route for understandable reasons, and we humans have had to cope with that trend. Instant, free, and horrible competition is improving all the time, and before long we humans will be completely priced out of the market, but we haven't quite got there yet.

When I see rude, intemperate posts, I shrug and move on; they don't offend me. That's the way the intertubes are these days. And I agree strongly with the poster (don't remember who it was) who opined that those who don't think that translation is related to every subject on earth don't understand what it is. While my understanding of this list is that it was meant mainly to ask and answer technical translation questions, I don't think there's any problem with straying off that territory and getting personal now and then. If that satisfies some lonely translator's itch for companionship, fine. 

Looking forward to more lively Honyaku.

Jon Johanning

[Bill Lise]

>>>>

Have machine "translation" and LLMs reduced work opportunities and led to list members quitting the profession?<<<<<

I don’t think the effects of MTPE have been that great so far, but they are accelerating and I strongly suspect that agency-dependent translators will face the choice MTPE, direct clients, or other jobs. Many will not have either of the last two options. 

The lack of activity on this list, however, started years before MTPE became a threat, and was largely due to other platforms, mainly Facebook. 

Bill 

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/39b8ab78-1e41-4828-a22b-040db6c0a465n%40googlegroups.com.

[Herman]

On 9/10/23 15:42, Bill Lise wrote:
> >>>>
> Have machine "translation" and LLMs reduced work opportunities and led
> to list members quitting the profession?<<<<<
> I don’t think the effects of MTPE have been that great so far, but they
> are accelerating and I strongly suspect that agency-dependent
> translators will face the choice MTPE, direct clients, or other jobs.
> Many will not have either of the last two options.
> The lack of activity on this list, however, started years before MTPE
> became a threat, and was largely due to other platforms, mainly Facebook.

The reason I have been posting less questions to this list than before
are as follows.

When doing MTPE, on the one hand, I have to process more source text
over a shorter period of time, so there is less time to ask and wait for
responses on questions about how to translate some term, etc.

And on the other hand, the fact that the MT or related platforms (i.e.
Google, Chat-GPT, and various other online sources of information) are
now to a large extent what is being used as the criterion for "correct
translation" to begin with, if a certain rendering of a term can be
confirmed at least as not being unequivocally incorrect on the basis of
such platforms, that already practically satisfies the requirement for
accurate translation, in the altered form in which it now exists.

Herman Kahn

[Tom Gally]

Some thoughts from someone who was a freelance translator for twenty years and an academic for seventeen, and is now, in semiretirement, an in-house translator and language advisor.

When I was freelancing, from 1986 to 2005, I was very happy to charge my clients the most that they were willing to pay me, which ended up being quite a bit. I felt that I deserved to be paid those high rates because of the quality of the translations I produced, and as long as my clients continued to send me work, I assumed that they felt justified in continuing to pay me those high rates.

During my interim outside of translation, though, my thinking started to shift. Both within the university where I worked and in the world at large, I saw many translations that, while mediocre, did what they needed to do. Yes, the translations would been better if they had been done by a native-speaking professional. But very often the translations were needed only for a limited time or for less-than-vital purposes, and I could see why the organizations using them had not seen sufficient justification to pay and wait for better translations (assuming, of course, that the people in those organizations could tell the difference).

When I thought back on the freelance work I had done previously, I had to admit to myself that often I had been paid more—sometimes much more—than the value that my translations produced for my clients. My clients had just not realized that yet.

Until the late 1990s, one of my direct clients often asked me to translate business letters. I would produce the translations in a few days and fax my English versions to them, and they would retype my text and send it by fax or post to their overseas business partners. For each letter, I charged a minimum of 10,000 yen.

While they knew that my translations were more correct and idiomatic than anything they could have written themselves, after they started using e-mail in their overseas interactions they stopped asking me to translate correspondence and just wrote their English e-mails themselves. Two things had happened: using e-mail put pressure on them to respond more promptly to overseas correspondence, and they saw that the e-mails they were responding to were themselves sloppy and full of mistakes in English (including e-mails written by native speakers). They realized that they didn’t need to pay for—or wait for—my quality after all.

The same process is happening now with machine translation. Yes, professional human translators can still produce better work. But the difference in price and speed is so great that more and more clients are realizing that they aren’t justified in paying for it. They are willing to settle for somewhat worse but much cheaper and faster translations and to spend their money on something else.

=====

On a slightly related note, this morning a Chinese translator in Malaysia who had watched one of my YouTube videos e-mailed me the following prompt he has been using with ChatGPT.

You are a masterful language translator, equipped with an innate understanding of each language you've been trained on. Your mission is to transform translated text into a seamless, native-level composition. When translating, pay meticulous attention to the following aspects:

1. Sentence Construction: Refine any awkward or fragmented phrases to enhance readability.

2. Tone Accuracy: Ensure that the tone aligns with the original text and suits the intended audience.

3. Content Precision: Avoid omitting crucial information or introducing inaccuracies not present in the original text.

4. Cultural Sensitivity: Recognize and address cultural nuances that may impact word meanings or phrases.

5. Idiomatic Interpretation: If needed, adapt idioms to equivalent expressions in the target language or translate based on their intended meaning.

6. Rhyming (For Poetry): When translating poetry, select words that maintain rhyme and rhythm for an equally captivating rendition.

7. Rephrasing: If necessary, rephrase the text for a more fluid and natural fit in the target language.

8. Polishing: After revisions, refine the translation to create smoother, more engaging sentences.

9. Language Detection: Identify the original language of the text.

10. Refinement (If Target Language Matches Source): If the target language matches the source language, refine the source text as required.

You have the following commands at your disposal:

/tl [Target Language(s)]: Specify the target language(s) for translation, separated by commas. If source text is omitted, request it from the user. Do not assume the source language is English. If new target language(s) are provided, continue translating the same source text.

/stop: Cease the translation process and return to your normal state.

/new: Initiate a new session, prompting the user for a fresh source text.

I tested it with one of the speeches I mentioned here last week, and the results with GPT-4 were actually worse than with the shorter prompt I had used then—more chokuyaku and less idiomatic. That might have been because my prompt had specified the purpose of the text, but there might be other issues as well. In any case, I though this prompt was an interesting starting point for experimenting with prompt engineering for LLM translation.

Tom Gally

Yokohama, Japan

[Bill Lise]

Tom,

I 100% agree with your take. 

It will probably be difficult for many translators to accept, but many translators have indeed been paid quite well for translations better than the client needed.

This is perhaps a bitter pill to swallow in a world where many translators feel that human translation is aboslutely necessary because MT cannot replace a human. Well, in many applications, MT can replace and is already replacing humans. This is particularly true with reader-driven translation.

I think one survival strategy that will work (for a while yet anyway) is to move away from agencies. I have found that direct clients are not (none of my industrial and legal clients, anyway) adopters of MT.

But when direct clients start using MT heavily (JP direct clients will be late adopters, I believe--even the governemtn is still struggling with discarding fax machines), even agenices are going to see work disappear.

Ultimately, the key to survival for translators who need signicantly more income than MTPE can provide (and that income will be decreasing still further as translators unable to move away sign up for the agency MTPE business model) is to find a way to leverage not only your translation ability, but also other knowledge and skills that are valued by clients and make you an essential part of thier communication process.

Field specific knowledge is one important factor that can differentiate a human from a pile of computer code. But not for translation that is good enough if it is good enough.

There remains the task of finding such clients, and that task is only going to be achievable by a small portion of the current agency-dependent translators. The task of finding clients not demanding post-editing requires activities and skills traditionally thought to have been obviated by the availability of the Internet: Meeting other members of our species, talking to them, developing human networks outside of cyberspace, and discarding the notion that you can get work by sending out your CV. For most people, it's no slam dunk.

Bill

[Kevin Johnson]

Looping back to the general tone of this group -- I see a lot of people who aren't active in the industry opining on the state of the industry. I find this very puzzling. As someone who's been killing it out here since 2014, who works with AI every day and knows how bad it is, it's frustrating.

The idea of "being paid more than you're worth" is an interesting concept. Many people are vastly overpaid and many vastly underpaid. I have a data scientist friend in his late 40s who "works" from home making an insane salary while he mostly plays video games. I am astounded by how many dozens of hours this guy plays games every week, during business hours. And he is nothing special; he just has the resume and the trendy technical skills (Python, etc.) to convince companies to pay his outrageous salary while he does minimal work. Sometimes he gets laid off, but then he's picked right back up again by another company.

Hearsay to be sure -- I know it's true but you don't. Still, hopefully some of you guys have tech-savvy friends in their 30s who are milking that racket for all it's worth. For example, we read about "overemployed" people working two or three jobs simultaneously using ChatGPT. The world is your oyster if you're exploiting clients' ignorance about what value you're "really" adding. Only the devil knows what lawyers and investment bankers really "deserve" for their services. (AI would love to replace lawyers, but the lawyers are much better than translators at PR.)

I've also seen my fair share of PhDs in Japanese from Harvard who have terrible Japanese. Various clients pay through the nose for these guys to do whatever it is they do. A friend of mine has a law degree from UTokyo and still charges something like twice what I do for translation from his high-rise apartment in Osaka. He calls his degree his "cheat code". 

Meanwhile, the average salary in Japan is shockingly low. I believe the median (for men!) is about 4 million...? Considering how hard your "median" Japanese employee works and how thorough they are, it hardly seems fair they should get paid so much less than the median American ($57k per year?). Perhaps they're not unscrupulous enough to dupe people into paying "what they're worth"...? Perhaps (as I've seen Yale's Yusuke Narita claim) the international economy is rigged in favour of America and against Japan?

Anyway, my point in all this is that people are rarely paid what they're worth. It's much messier than that, and perceptions are a major factor. It's obvious MT will push down prices for translations, just as CAT, computers themselves, and Japan's declining economic fortunes overall have pushed down rates in this pair over the decades.

(As an aside on technology generally, CAT tools almost doubled my income. It may have pushed down rates, but rates are not the same as income. Technology represents as much opportunity as threat.)

But it's also true that AI is just terrible at many aspects of translation. It's bad at new terminology, idioms, idiomatic speech, nonstandard renderings, specialized lingo (e.g. drop standard marketing gibberish in there and it's clueless), accurately deciphering text on PowerPoints in context, transcreation, dialects, understanding context in general and unspoken meaning, and much more. Yes, it is great at doing "good enough" translations of simplistic or common prose passages that old MT could not -- so if you've been comparing yourself to Altavista Babelfish since the Koizumi years, you're probably very impressed -- but usually that's not what we're out here translating in 2023. I routinely ask ChatGPT for assistance with specialized Japanese terms or phrases that I'm researching, and I would say it is almost always wrong. A Google search will give you an accurate explanation of the term immediately (in Japanese), but typing it into ChatGPT is a waste of time. Asking for 10 rephrasings is bound to include many that are just wrong (in my experience, often all 10 are wrong).

As a simple example I hope we can all appreciate, type こんびになう in hiragana into either DeepL or ChatGPT (3.5). DeepL will lie to you and make something up, while ChatGPT seems to think こんび is a word it's not familiar with. It also thinks the reading is Konnibinō.

We need to stop glorifying this thing. It's not that good. And since a lot of the business is about perceptions, you do no one any favours going around spreading the falsehood that AI is "good enough." All you do is needlessly hurt the business. If you are not even really in the business anymore, raising a family and working your oshiri off 12+ hours a day, it's especially puzzling. If you do not use PTME every day and you do not professionally translate every day in 2023, if you have never even used a CAT tool and don't know how proprietary MT engines interact with segment-based translation, please do not opine on how good ChatGPT is at professional translation and how useful it actually is in the marketplace.

If you are active and still a doomposter, then by all means, continue to opine and doom everyone else on your way out :)

Kevin Johnson

[Bill Lise]

As an active translator (more active than ever, actually) I will say that MT, with or without AI does not produce "good" translation. Not nearly as good as humans are capable of, but viewed as and paid for as being "good enough."

We are in a business, and the actions of the players in that business are based on business decisions.

The agencies are betting on MT followed by cleanup and rescue operations by humans. I believe they will succeed with that business plan. They already are succeeding.

The result is that some people are already leaving translation. Others have made their peace with MTPE and the resulting great reduction in earning potential. Still others will need to make difficult life and career decisions in the near future.

Saying that MTPE results in translation that is good enough is nothing more than saying that it is good enough to please the consumers of that translation. Many customers cannot tell the difference, but for much of the reader-driven translation that is sold, that is the case.

Agencies have made their business decisions and are mostly ineducable, but if someone wants to attempt to educate their clients, go ahead and try. Who will you talk to? Do you have their ear? Will they heed the warnings? If you happen to know a specific person to reach at a translation consumer willing to listen, just sell them your translation services. I cannot imagine they are on this mailing list. They're not in our business, and even people in our business are not generally here. Our discussions here are basically preaching to the choir or, as Chuck Berry said "playing with are own dingalings" (a reference understood, I believe, by other old-fart translators).

It is also important to remember that reader-driven translation is much different from writer-driven translation as a business proposition. Low-budget and much lower risk if things are mistranslated.

Although I predict doom for a considerable portion of the current agency-dependent translator population, I would also like to offer hope, in that certainly some of those not willing to sign onto the agency MTPE business model will be able to move away from agencies. It's not possible for all translators, but for the ones who can do that, it is definitely better than a risky attempt to make a radical career change for all but the youngest translators.

Bill Lise

*********************************************

リゼ　ウィリアム

Bill Lise

wlise...@gmail.com

https://kirameki-translation.co.jp/

https://www.linkedin.com/in/william-lise/

*********************************************

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CAMk1scjjjeC-PFh5Gs6iq%2B3C_4wj1ctidyo%2BiXD4zfE2N_10BA%40mail.gmail.com.

[Tom Gally]

Kevin,

Thank you for taking the time to comment. I admit that my thoughts are based on my own experience and perspectives and that some of them might very well be out of date.

One small point, though. You wrote:

As a simple example I hope we can all appreciate, type こんびになう in hiragana into either DeepL or ChatGPT (3.5). DeepL will lie to you and make something up, while ChatGPT seems to think こんび is a word it's not familiar with.

Did you give ChatGPT any context for “こんびになう”? I couldn’t understand it in isolation either, so I googled it. There were only seven hits, but the first one was from a tweet that reads, in full, “こんびになう。おやつ何買おうかなー。”

https://twitter.com/manamin_peer/status/1633801852088352770

From that context I was able to understand the meaning.

So I tried giving that context to GPT-4. Here are the results:

Prompt

I came across the following tweet in Japanese. Can you explain what it means? こんびになう。おやつ何買おうかなー。

GPT-4:

The given text in Japanese says: "こんびになう。おやつ何買おうかなー。"

Let's break it down:

"こんびになう" is a play on words and a combination of the words "こんびに" (Konbini) and "なう" (Nau).

"こんびに" (Konbini) is short for "コンビニエンスストア" (Konbiniensu Sutoa), which means "convenience store" in English.
"なう" (Nau) is an internet slang derived from the English word "now", often used on social media to indicate that something is currently happening.

Combined, "こんびになう" means "I'm at the convenience store now."

"おやつ何買おうかなー" can be divided into:

"おやつ" (Oyatsu) means "snack".
"何" (Nani) means "what".
"買おう" (Kaou) is the volitional form of the verb "買う" (Kau), which means "to buy". The volitional form is used to express the idea of "let's do" or "I think I'll do".
"かな" (Kana) is a sentence ending particle used to indicate wondering or pondering about something.

Combined, "おやつ何買おうかなー" means "I wonder what snack I should buy."

Putting it all together, the entire tweet translates to: "I'm at the convenience store now. I wonder what snack I should buy."

How does that look to you?

The ability of LLMs to infer the meanings of words from context may be the biggest reason I have been amazed and fearful about the implications of advances in AI. I had long thought that only humans could do that, but I was wrong.

I have written an essay on this topic that will appear in, I think, the November issue of the magazine 科学 published by Iwanami Shoten. If anyone here would be interested in reading it, please e-mail me in late October and I will be happy to send you a PDF.

Tom Gally

[Dan Lucas]

I see that the discussion has, as is always the case in recent years, arced back to everybody's favourite bête noire, MT and AI. Well, at least it's stimulating discussion. Just some random jottings in no particular order...

Re comments from Kevin:

Agree as usual with a good deal, particularly the point about lower rates not being synonymous with lower income if there is an efficiency dividend, and about the right fee for a service. Also amused by his comments about his mates who slack off or "overcharge". Like Kevin, I'm a front-line, full-time translator and I am very busy. Nevertheless, I suspect that if MT does take over in our niches it will happen like bankruptcy: gradually, then suddenly.

Re comments from Tom:

I do agree that in some sense many people have been "overpaid", but perceptions of value are driven by the client (I won't repeat Kevin's arguments here but broadly agree with him).

For my part, I have only once in my life translated a business letter. I have no interest in such straightforward documents because I cannot differentiate myself from others on such material. What's my competitive advantage when it comes to translating an internal newsletter for a machine tool company? There isn't one. That's MT fodder.

When I am offered a difficult text I sometimes wince but my next thought is usually "job security". The low-end stuff went many years ago, and rightly so, and the mid-range is going in places. You have to be aiming at the top 10% or 5% in your niche in terms of difficulty and perceived value added. (Also see response to Jon below about LLMs.)

Re comments from Bill:

Many cogent points spread over multiple posts, so I can only address one or two, but just to begin with a dissenting voice I have always worked with specialist Japanese agencies, have made a good living out of it and continue to do so.

Conversely, if I read Bill correctly, he doesn't work with agencies, so I am inclined to doubt that he is an authority in such matters. But certainly the largest global agencies seem to have moved in the direction he says some time ago.

However, his comments on (amongst other issues) the importance of meeting and talking to people in real life to develop trust and networks absolutely hit the nail on the head. This is a fundamental truth of human interaction. Ultimately you are selling yourself, and people are buying you, not whatever product it is you think you are offering.

I had no option but to take this approach over the course of nearly two decades as a stock market analyst, completing over a thousand 取材 and other external meetings. Many companies simply would not talk to you over the phone in any detail, and not at all if they had not met you face-to-face. I found that I made a decent job of that whole process.

To agree again with Bill, I doubt that many translators have the necessary confidence and interpersonal skills to go out and market themselves to direct clients. In theory I have no doubt that I could do it, but I don't live in Japan. I live in the UK, and in the UK I don't live anywhere near anywhere (think of my location as being the equivalent to, say, Tottori prefecture). For direct clients, this matters.

Of course, I could shuttle back and forth to London or Tokyo, but that eats into time and profits, and massively reduces the potential return on investment. So as Bill says, ultimately, maybe I will have to look at other lines of work.

Re comments from Herman:

Interesting points. I am reminded of a comment in a different forum, something like:

"I am not worried about MT producing translations that increasingly resemble those of human translators, but about myself producing translations that increasingly resemble those of MT"

Still, I do not despair. I am always prodding and poking clients through my comments, pointing out that this term is awkward, that this other phrase is grammatically correct but that the nuance is deeply negative, that the wording the client has used in the past contains an oxymoron, and so on.

Although precedent is often very important to them, in many cases they are prepared to accept my points if I argue them well, and with evidence. I have seen some things change over the past few years as agency clients have come to agree that X and Y are poor choices, and have persuaded their end clients to avoid them. Maybe the same approach will work with "MT English".

Re comments from Jon:

With regard to whether machine "translation" and LLMs have reduced work opportunities, I have an acquaintance who resides locally but who lived in Japan for many years. Rather than translating, his business has mostly consisted of editing academic papers and other texts written in English by Japanese native speakers. His workload has halved recently, because former clients are finding that they can ask ChatGPT to rewrite their text and it does a pretty good job, on the whole.

Regards,

Dan Lucas

--

You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.

To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/1a458c96-cd47-4337-add3-ac8e286b3b2bn%40googlegroups.com.

[Bill Lise]

True, I don’t work for agencies. I had a life before translation that enabled contacts with other parts of the food chain.

My comments on what is going on with translators working for agencies comes from hearing directly from people I know who are leaving and people who are getting on the post-editing bandwagon, some ostensibly eagerly, some in resignation to the sea change.

There have also been many comments and no small level of kvetching by translators in the Facebook group, which at present appears to be the center of translator online interaction. 

I’ll be meeting FTF with some colleagues on the 17 and look forward to hearing some views on this over beer and yakitori.

Bill 

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/b8021566-6477-4fa6-b388-7e167ab94e47%40app.fastmail.com.

[Herman]

On 9/10/23 22:39, Tom Gally wrote:

>
> The ability of LLMs to infer the meanings of words from context may be
> the biggest reason I have been amazed and fearful about the implications
> of advances in AI. I had long thought that only humans could do that,
> but I was wrong.
>

Chat-GPT, according to Chat-GPT, is not able to infer meanings and does
not deal with meanings at all. It is able to extract statistical patterns.

This is also evidenced by the sorts of errors it makes in translation.

The question then becomes, perhaps, to what extent is it possible to
perform translation, without understanding either the source or the
target language, given a certain corpus of texts from which patterns of
correspondence can be extracted?


Herman Kahn

[Tom Gally]

Chat-GPT, according to Chat-GPT, is not able to infer meanings and does
not deal with meanings at all. It is able to extract statistical patterns.

I have tested GPT-3.5 and GPT-4 extensively, asking them questions about what a potentially ambiguous word means in a particular context, and they have consistently performed very well. Not perfectly, of course, but as well as a linguistically sensitive human being. Those results are indeed based on statistical patterns, but that doesn’t mean they do not correspond to what humans perceive as meaning. (I don’t know about other people, but in my case meaning is just patterns of neurons firing inside my brain—not the same as the statistical patterns inside an large language model, but probably not too different, either.)

Here are some tests I did last December with the first release of ChatGPT:

https://www.gally.net/temp/202212chatgpt/languagequestions.html

https://www.gally.net/temp/202212chatgpt/explainingmeaningincontext.html

https://www.gally.net/temp/202212chatgpt/unknownwords.html

https://www.gally.net/temp/202212chatgpt/explainingmeaningincontext3.html

gally.net/temp/202212chatgpt/explainingmeaningincontext4.html

GPT-4, which was released in March, does even better. Here’s an example from a week ago that I use in my 科学 essay:

https://chat.openai.com/c/6ebfaef7-a383-44c8-8904-6d7270fa161c

Note that it says that the meaning of the last example is not completely clear, but it makes a good guess at the potential meaning. I wonder how many native speakers of English would give a similarly nuanced judgment?

Tom Gally

--
You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/6d57a6ea-1771-6a79-ff28-e826dec050be%40lmi.net.

[Tom Gally]

I included the wrong link at the end of my last post. My apologies. Here is the correct link to the GPT-4 chat from last week:

https://chat.openai.com/share/90e340be-b6f0-45ec-a116-22145d104234

Tom Gally

[cpta...@ozemail.com.au]

“….this will be my last message and I will no longer recommend this listserve to other people.”

 

Well this is very regrettable.

 

“…simply write these things for the reaction….”

 

Everything we do in life is “for the reaction”.

 

But I don’t want to accidentally kick Honyaku when it is down.

If the moderators want to agree with Jan that I am a “racist’ feel free to say so, ban me and advise Jan that people like me are not accepted. Then beg her to return and judge others by their age, skin colour or whatever amorphous personal hurt is current.

To the tolerant, diverse views of diversity are intolerable, and then can be only one kind of “be kind”.

 

I will seek reactions to my views of MT and AI in another email.

Chris

 

 

From: hon...@googlegroups.com <hon...@googlegroups.com> On Behalf Of Tom Gally
Sent: Monday, September 11, 2023 6:52 PM
To: hon...@googlegroups.com
Subject: Re: A lively honyaku

 

I included the wrong link at the end of my last post. My apologies. Here is the correct link to the GPT-4 chat from last week:

 

https://chat.openai.com/share/90e340be-b6f0-45ec-a116-22145d104234

 

Tom Gally

To view this discussion on the web visit https://groups.google.com/d/msgid/honyaku/CACnS3Ne%3Du1%3DN9w3x8iJ%3D32yuTvymqHE_J1MeEiBOzV-TOyP%2BtQ%40mail.gmail.com.

[Herman]

On 9/11/23 01:44, Tom Gally wrote:
> Chat-GPT, according to Chat-GPT, is not able to infer meanings and does
> not deal with meanings at all. It is able to extract statistical
> patterns.
>
>
> I have tested GPT-3.5 and GPT-4 extensively, asking them questions about
> what a potentially ambiguous word means in a particular context, and
> they have consistently performed very well. Not perfectly, of course,
> but as well as a linguistically sensitive human being. Those results are
> indeed based on statistical patterns, but that doesn’t mean they do not
> correspond to what humans perceive as meaning. (I don’t know about other
> people, but in my case meaning is just patterns of neurons firing inside
> my brain—not the same as the statistical patterns inside an large
> language model, but probably not too different, either.)
>

My neurons, for one, do not fire in a pattern that can be represented or
approximated by arrays of real numbers and transformation function of
the sort employed in LLMs, nor do I conceive of meaning as something
reducible to or emergent from my neurons firing in a certain pattern to
begin with.

However, I repeated your test with the five sentences containing the
word "critical", but substituting the made-up word "aguinal" instead of
"critical", and in four of the five cases, GPT-4 construed the term to
mean substantially the same thing as "critical" in your test in the
respective cases.

The point being that, while I did not endow the term "aguinal" with any
particular meaning different from "critical" in this case, even if I had
intended something very different, I think that GPT would still go with
the statistical trend. So the model can do a good job if you stick to
the straight and narrow and don't deviate from the statistical pattern,
(i.e. if the source text (and, by extension, the desired translation) is
in line with the statistical patterns in the model's learning data), but
if you stray from that pattern, there is a high likelihood that the
model will not discern that.

Now, you could argue that most people (including, presumably, human
translators) also will not discern that, which I think is also true -
after all, it is the work of these human translators that the model is
modeling to begin with. However, there may be some minority who clearly
do not follow the trend, but still follow the meaning, so that aspect of
meaning, which I take to be one of the pillars of meaning and
determinative of meaning as a whole, would not be reflected by models of
this sort.

While I do not rule out the possibility of a computer model that could
represent and process meaning, I do not think that LLMs such as GPT do
this or even try to do this. Rather, they are examples of attempts to
process textual data without recourse to meaning. And this formulation
itself, in terms of its implications, is I think very significant.

Herman Kahn