New issue of Translating and Interpreting

Ignacio Garcia has just sent out the latest issue of The International Journal for Translating and Interpreting Research, including articles on speech recognition in translator training (Dragsted, Mees & Hansen), translation memory-mediated environments (Mesa-Lao), legal interpreters in Ireland (Phelan) and a quantitative study on clear English for Translation (Burns and Kim). Personally, I’m most excited about Pym‘s What technology does to translating:

Abstract: The relation between technology and translating is part of the wider question of what technology does to language. It is now a key question because new translation technologies such as translation memories, data-based machine translation, and collaborative translation management systems, far from being merely added tools, are altering the very nature of the translator’s cognitive activity, social relations, and professional standing. Here we argue that technologies first affect memory capacity in such a way that the paradigmatic is imposed more frequently on the syntagmatic. It follows that the translating activity is enhanced in its generative moment, yet potentially retarded in the moment of selection, where the values of intuition and text flow become difficult to recuperate. The redeeming grace of new technologies may nevertheless lie in new modes of opening translation to the space of volunteer translation, where humanizing dialogue can enter the internal dimension of translation decisions. The regime of the paradigmatic may thus be embedded in new modes of social exchange, where translation becomes one of the five basic language skills.

My main interest, of course, is in asking those questions that Pym potentially doesn’t consider. I mean this without malice, but I do feel that the academy is carefully cotton-wooled itself from the more interesting ideas that have come from the last century. In particular, I’d be looking to ask questions like how does détournement affect his “humanizing dialogue” – for instance, in the realms of crowd-sourcedillicit subtitling.

I’ve yet to read the article, but am looking forward to it.

Algorithm detects tweeter’s gender

/. is reporting that “linguistic researchers have developed an algorithm that can predict the gender of a tweeter based solely on the 140 characters they choose to tweet”:

The research is based on the idea that women use language differently than men. ‘The mere fact of a tweet containing an exclamation mark or a smiley face meant that odds were a woman was tweeting, for instance,’ reports David Zax. Other research corroborates these findings, finding that women tend to use emoticons, abbreviations, repeated letters and expressions of affection more than men and linguists have also developed a list of gender-skewed words used more often by women including love, ha-ha, cute, omg, yay, hahaha, happy, girl, hair, lol, hubby, and chocolate. Remarkably, even when only provided with one tweet, the program could correctly identify gender 65.9% of the time. (PDF). Depending on how successful the program is proven to be, it could be used for ad-targeting, or for socio-linguistic research.”

Literature in pidgin

Boingboing has a link up to our very own ABC’s recording of Three Little Pigs in New Guinea Pidgin:

Talking about Pidgin on radio prompted Ralph Newton to send in a copy of Tripela Liklik Pik (Three Little Pigs) he’s had since he spent time in PNG in the 60’s. Click on the related audio link below if you’d like to hear what Pidgin sounds like.

The back cover of the record says: “This unique story of the Three Little Pigs was translated into Pidgin and adapted to a Melanesian setting by The Reverend Paul Freyberg of the Lutheran mission at Madang. Mr Freyberg was the Chief Translator of the Nupela Testamen – the New Testament in Pidgin. The story was broadcast by Superintendent Mike Thomas in the ABC’s Daily Learning Pidgin Series”.

Also mentioned is a pidgin version of Macbeth, known as Makbed, which I *wish* I could find an mp3 of.

Social Networking

I’ve not been very social recently – at least, not on the blog. It’s been a busy month of school holidays, tax returns and server migrations. I’m now providing a better service for my small client base and have enjoyed learning new technologies and thinking about what I have to do next to make it better still.

In the meantime, the world hasn’t stopped. The Murdoch empire’s underbelly has been exposed, Norway has experienced a shocking act of terror and Google has released a new social networking system – each of which has produced many a column inch.

I wont be talking on Murdoch or Norway – far smarter people than me have dealt with it effectively elsewhere – but I will speak on GooglePlus. I’m not going to explain what it is, or even go into much depth about what it does differently to FB, MySpace or Twitter.

I have an account, but my re-entry to SN wasn’t easy. After leaving the FB space almost a year ago, I discovered that I’d quite enjoyed the time without my inbox being filled with notifications about a long trail of friends, many of whom I’ve not been in contact with since leaving FB.  at the Guardian has written in Friendship: why social networks are too crowded to get close:

Also, as far as I can tell, each additional node on an online friendship network reduces the significance of the connections that came before. It’s impossible to keep up. I can’t identify what’s important anymore, and so rather than interacting, I just offload. That doesn’t help forge friendships; it becomes a one-way communication platform. You’d not say you’re friends with the people on your telly, so how is my stream-of-consciousness twitterfeed any different?

While this rings true for me, I don’t think G+ (or any SN) is all bad. In fact, G+ is actually doing SN better than others have previously, although some decisions have been terrible – the insistance on a real name policy is flawed:

[N]ames are not simple. They are not constant, unique identifiers. Danny O’Brien lists some common falsehoods about names:

[That] people’s names are assigned at birth; people’s names won’t contain any well-known English swearwords; and there is one way to spell someone’s name. But perhaps the most prevalent misconception that programmers cling to about names is that you have a single definitive name that stands above all the others – and that single name matches your legal identity.

Google also prohibits names that mix languages, but the idea that you have one name and that it’s in one language is manifestly wrong, as illustrated by CopyLion on the Gmail Help Forum

Hong Kong is a former British Colony and we use English names and nicknames far more than Chinese names and Pinyin of that.
[…]
How Hong Kong people’s name on their ID card typically looks like is like this:
CHAN, Tai Man

In Google’s favour, you can actually vote on changes to the system, here is one titled “Google+ should allow pseudonyms. This can be important for individuals who are known by a specific Alias or for people from politically unstable countries without freedom of speech.”

In the remainder of the pro column, Andreas Kluth‘s The sociological breakthrough of Google+ makes some very good points about discretion:

Because Facebook is fundamentally (=unalterably) indiscreet.

And it is fundamentally indiscreet because it is architecturally indiscrete. (Forgive me that word play.) Meaning: you cannot distinguish easily between different degrees of intimacy among the people in your social graph. The various relationships are not discrete, not separate.

Tell me: In real life, how often do you walk up to somebody and request to be “friends”, then begin “sharing” pictures of your naked baby?

Which he differentiates against G+:

Google+’s crucial innovation (among many others existing or planned) is Circles. You can make as many of them as you like. They can contain 1 person, 2 people, the Dunbar number, or the entire web. Because there are things you want to share with just one person, or with 2, or with lots, or with everybody (as on WordPress).

Ergo: Discrete → discreet

You also don’t have to ask anybody to be your “friend”. Nor do you have to reply to anybody’s “friend request”. You simple put people into the discrete/discreet spheres they already inhabit in your life.

I think this is an important point – it shows that those that provide services like these are learning all the time how to bridge the gap between what we can imagine the web can be, at it’s best and what we actually have. Meanwhile, Reilly has an interesting, if somewhat prematurely overoptimistic in my mind, article on why this isn’t G+ v FB but does go into what FB did wrong:

Facebook’s chief flaw is that is a closed platform. Facebook does not want to be the web. It would like to draw web citizens into itself, so it plays on the web, but in terms that leave no room for doubt where the power lies. Content items in Facebook do not have a URI, so by definition can never be part of the broader web. If you want to use Facebook’s social layer, you must be part of and subject to the Facebook platform.

Additionally, there are issues with the symmetry of Facebook’s friending model: it just doesn’t model real life situations.

In the meantime, I’ve seen many references to the quote if you’re not paying for it; you’re the product – not a new idea – true of newspapers, true of television, now also true of the internet. (Interestingly, the revitalised discussion of rights to privacy in regards to the Murdoch scandal, bleed into this area – it will be interesting to see what comes of it all).

If that is a real concern, and I think that it’s a reasonable one, there are options available. Google have already started suspending accounts (in fairness, at least one was apparently due to the gaming of adwords has since been retracted) and if you are using any other Google services, like Flickr, Gmail, GDocs and GCalendar – it can be disastrous. A Slashdot commentor gives some good advise for protecting yourself including the Google Takeout:

a Data Liberation platform that makes escaping from Google products as easy as possible.
Takeout lets you take your data out of multiple Google products in one fell swoop. Moreover, you’ll find that all your data is in portable and open formats‚ so it’s easy to import to other services quickly.

I don’t know if I’ll be going so far as Google Takeout, but I’ll definitely be implementing an offline backup of my Gmail – I’ve not used another system for over six years, it contains everything I’ve done, and it doesn’t use Thunderbird, which hasn’t impressed the last couple of times I’ve used it.