TAG | regulation
A few days ago I was congratulating Graca Machel (current wife of Nelson Mandela) for being the only woman ever to have been the First Lady of two different states. I then realised that a) the bus was moving very slowly, and b) I didn’t know much about her first husband, the socialist post-independence President of Mozambique. Enter Wikipedia, pursued by a bear. After reaching the climax (death by plane crash – at the hands of either inebriated soviet pilots or merciless Apartheid agents), I noticed that the article could be read in *31* alternative languages. Alongside the major ‘international’ languages (by which I mean, widely spoken as second languages or studied outside their regions of origin, the article is also available (albeit in greater brevity) in Belarusian (7.6m native speakers), Breton (206k), Catalan (11.5m), Estonian (1m), Georgian (7m), Latin (0), Lithuanian (3.2m), Mongol (5.7m), Occitan (2m), Venetan (2m).
On scanning the list, my initial sentiment was one of embarrassment – embarrassment that I didn’t know that Belorussia or Venice even had their own tongues, or that Occitan was even a thing anymore (it appears to have more native speakers than than all six Celtic languages combined).
Emerging from this introspective shock, I noticed that the article only appeared in one language of African ‘origin’ (not the sturdiest of criteria, but here I’m excluding French, Portuguese, English and Arabic) – Ido. This I found bemusing, as I was pretty sure that Ido was West African (ergo, pretty far from Mozambique). Once again smacked down by Wikipedia, I soon discovered that I was thinking of Igbo (spoken by 24m in Nigeria and Equatorial Guinea); Ido actually appears to be a descendant of Esperanto, and has a paltry 100-200 ‘users’.
And so, I was kind of stunned that an article about one of sub-Saharan Africa’s great independence leaders is available in a range of languages demographically-minor European languages (almost all speakers of which are multilingual), but not a single non-European language of Mozambique, or indeed of sub-Saharan Africa. According to a 2007 survey, Emakhuwa is the mother tongue of around a quarter of the Mozambican population, while only 10% claim Portuguese as their first language (although it is the second language of a further 40%).
Wikipedia handily details all 286 languages in which the page on Samora Machel is available, ordering them by the number of articles for which they are used. Incredibly, only seven ‘African’ languages are in the top 200: Yoruba (75th, with ca. 30k pages); Afrikaans (82nd, 26k); Swahili (84th, 25k); Amharic (112th, 12k); Somali (179th, 2.5k). ; Lingala (194th, 2k); and Kinyarwanda (200th, 1.8k). A further 31 African languages have a ‘wiki-presence’.
It makes sense; if you need to use the web to publicise a product, service or opinion, using Tswana, Kikuyu or Ewe makes little sense. It goes without saying that the global reach of the internet forces suppliers to market their wares/thoughts to a wider audience than they would in previously more localised markets. Uganda alone has over forty languages; although many within the same linguistic group may be mutually intelligible to a degree, the much wider comprehension of English makes it a much more dependable choice for anyone looking to use the web for intranational purposes (let alone international).
It’s not feasible to hope that individuals and businesses in African countries devote time to the creation and maintenance of web resources in languages other than (and possibly in addition to) those which will allow them to achieve their key economic or philosophical goals. It would actually be pretty paternalistic. But it is pertinent to ask if, as the web becomes an increasingly important mode of communication in ‘developing’ countries, can indigenous languages survive as means of verbal communication while being relegated to digital redundancy? UNESCO estimates that 90% of individuals in developing areas have no access to broadband, but a raft of initiatives (e.g. Inveneo’s BB4G) employing new business models and cheap technologies are changing the connectivity landscape at pace. Avanti’s HYLAS 2 satellite, launched in August 2012, has reduced dependence for millions in eastern and southern Africa on unreliable undersea cables. While cost may still be prohibitive to many, there’s an expectation this will decrease over time. As societies in Africa become increasing engaged in digital communications, how will the position of indigenous languages in the non-digital sphere develop?
It may even be possible to correlate the rise of the internet with the recently declining status of indigenous languages. Between the 1997 and 2007 Mozambican censuses, there was (apparently) a significant increase in the number of people claiming portuguese as their mother tongue to 10%. 42.9% of the inhabitants of the capital Maputo held the post-colonial language as their first language in 2007, although I couldn’t find out what this rose from. It would be interesting to see if it’s possible to plot sociolinguistic identity against the growth in internet access, although undoubtedly other factors must have major roles to play (tourism, demographic shift, increased literacy). If anyone knows of any relevant literature, forward it on!
Millions of people sleep at night completely unaware that personalised search has steadily been revolutionising the way they see the web and, as with most Google led initiatives, there’s essentially nothing we can do about it…or is there?
Imagine the internet as a vast department store filled with everything you could possibly think of and many things you’d rather not. Google’s place in this department store is the plucky store attendant; tell them what you’re looking for and they’ll gladly guide you to the relevant bits of the store. In the carefree days before personalised search, Google would guide everyone searching for ‘shoes’ to the same shop filled with only the most relevant items, presumably shoes. Now this store attendant takes us to our own bespoke storefront filled with shoes in exactly my size and taste; some utopian retail fantasy where the bricks and mortar are data culled from my past shopping excursions and those of my friends. It’s a favourable analogy: this notion of personalisation has been the death of the retail high street. In the online world where everything is freely available in shiny web 2.0 style, the consumer and search user expects, nay demands, to get exactly what they want.
But the façade of user control is a thinly veiled one. The problem with the covert landing of personalised web over the past few years is just this – it’s entirely justifiable from a user experience point of view. Hate them as you will, but the Panda and Penguin updates did actually provide better results for the casual searcher; this has always been Google’s prerogative despite the indignant gnatter of the SEO’s keyboard. This new UX incentive seems a little twisted though. Clearly everyone likes what they like so there is considerable justification in throwing up related items again and again in search results. But we may also like things we don’t know we like and more importantly, things Google doesn’t know we like. Perhaps I don’t want the same Italian restaurant in my area, but even after the food poisoning subsides the search history remains. I want to holiday somewhere new this year, far away from all my friends and everything they like and generally everyone else on the web, will Google let me? What if I want to boldly go where my search history has never taken me before?
Internet searchers have responded to the increasing relevancy of Google’s search results by making it their shop attendant of choice (alas poor Jeeves, I knew ye not), but personalised search have moved away from this key metric. The notion seems to be that by showing us results we’ve responded to before before, Google is second guessing our future preferences, presumably to make the choice easier, quicker or remove it altogether. But my search history is a terrible approximation of who I am and what I want and thus what is relevant to me. What if I’m on a public computer, my friend’s browser, his Gmail account, what if I mistyped those saucy search terms and never want Google suggesting them at work again? The illusion is that I have already chosen these results through my past search activity, that I am in control here – the reality is that Google’s algorithms are in control.
So is personalised search better in any sense? The pedantic answer is that it depends what you mean by ‘search’; Google, after all, would be the first to concede that not all searches and searchers are the same. Whether any of this bothers you or not will probably depend on whether you’re the semi-mythical user who always knows exactly what they want, or, more likely, if you’re just conducting a basic informational search. But to fall in the other camp you don’t even have to construct leftist arguments about equality and freedom of online information to all, you just have to posses that very human characteristic of mutability.
One site that seems to recognise this is Match.com. Their occasional suggestions of people who don’t fit your ‘type’ recognise the fascist half-truth that the people probably don’t know what’s good for them. ‘I am not just what I search’; a new kind of social rallying-cry against the algorithmic oppressors. Don’t shout it from the streets, rage against the search engine my friends. Turn off personalised search now and rediscover the joy of finding what you weren’t looking for.
The standout moment at last month’s eG8 summit in Paris saw Nicolas Sarkozy offer a foreboding warning that the internet must not become a ‘parallel universe without rules’ – only days before David Cameron had been at pains to distance himself from the idea of state regulation of the internet. But why is it that the same morality and rules of law that we defend culturally are seemingly so inapplicable to human interaction over the net? The question is one which is rapidly forcing internet moguls like Mark Zuckerberg, who also addressed the eG8 summit, straight into the ring with political leaders.
It’s clearly an issue for governments and the internet industry to consider. Responsibility for regulating the web has for too long seemed a question impossibly gargantuan, perhaps too hopelessly multifaceted to be properly addressed by heads of state. A more accessible dialogue on what law is needed in cyberspace might have prevented the abuse of its liberal merits by tabloid newspapers in privacy scandals such as the failure of Ryan Giggs’ gagging order, whereby papers stake claim to a better representation of our rights as net-users than law courts do. As with the Space Race and contested rights to Deep Sea Oil Reserves in the antarctic before it, the internet seems to lack the clear geographical or institutional boundaries which would validate an open discussion on its regulation in national or global fora.
Interestingly, Rupert Murdoch was amongst the crowd who received Sarkozy’s assertion that governments must not allow the internet to remain unchecked. Looking at British politics (almost unavoidably through the window of a Murdoch-owned medium), it is hard to argue against any regulation of the internet. Just as parliament and the English courts are sometimes made to look irrelevant by the power of Murdoch’s media and the twitterati masses, Mark Zuckberg also presented the case for an entirely unregulated global space.
Zuckerberg said: “I’m happy to play any role they [the people] ask me to play… the internet is really a powerful force for giving people a voice.” In fact Zuckerberg openly undermined Sarkozy’s opinion througout the eG8, adding: “People tell me: ‘It’s great you played such a big role in the Arab spring, but it’s also kind of scary because you enable all this sharing and collect information on people…But it’s hard to have one without the other. You can’t isolate some things you like about the internet, and control other things you don’t.”
First things first: ‘What is ad retargeting?’. Simply put, it’s advertising targeted specifically at you, based on things you have already shown an interest in, but not bought. Say, for example, you were mulling over a lovely pair of socks on your favourite sock retailer’s website but navigated away without purchasing said pair of socks. The chances are, you will suddenly notice adverts appearing on websites you visit later on that day as if you were being hounded by some kind of relentless electronic sock salesman (an RESS, as it’s known in the industry). Coincidence? No. You’ve been retargeted.
This sort of advertising is set to start happening a lot more in the coming year. Some of the major retailers are investing heavily in ad retargeting, making it one of their key online marketing strategies for 2011. The reason they’re switching investment to this area: it works. All Saints (the clothing retailer, not the girl band), say they generated a return of £21 for every £1 spent on retargeting ads in the last two months of 2010. Struq, a specialist retargeting company, claim they are generating conversions of up to 640% for their top ten clients. So, you can clearly see why it’s appealing to online retailers. But how is sitting with the consumers?
Joseph Turow, a professor from the University of Pennsylvania and a specialist in online advertising, thinks that people really don’t like it once they learn how it’s happening. Whereas to others, perhaps, it’s not quite as bad as full-on behavioural advertising because you know why these adverts keep following you around. But what if you don’t want your shopping interests to follow you around the internet? You left the site because you decided not to buy the product, now it’s appearing on every other site just to torment you. Worse yet, you might find products being targeted at people who you share your computer with, products that you really don’t want to be brought to their attention. There are some companies, such as Criteo, that provide an opt-out option at the bottom of their retargeted ads but unfortunately this is not a universal principle followed by all advertisers. A better solution may be appearing as browser firms are working on integrated ‘do not track’ systems, like the recently released Keep My Opt Out extension for Google Chrome which keeps opt-out settings even if your cookies get cleared.
The rise in retargeted advertising is not going unnoticed. The Institute of Practitioners in Advertising has called for immediate research and the formulation of best practice around its use. Although this is borne of a fear that there will eventually be a backlash from unhappy consumers, it is a positive nevertheless. This is an issue our friends at the Internet Advertising Bureau are tackling by including it in their EU self-regulatory good practice framework. As highlighted in our preview of the 2011, online privacy is a hot topic at the moment. If the issue of retargeted advertising is not developed with consumer control and transparency in mind, there could well be an eventual uproar from consumers. What’s more, if left to develop freely, along with personalised behavioural advertising, what’s next? Personalised biometric advertising, Minority Report-style?
Consumers may not be totally adverse to retargeted advertising: there are circumstances where it could be truly useful – where it’s used not just to readvertise something to you but if it’s remarketed in a different way. That pair of socks you decided not to purchase may have been because you thought the price was too high so you went to look somewhere else. If, later on, an advert for those socks reappeared on a different site telling you they had been reduced by 50%, you might be extremely happy and praise the ingenuity of those clever retargeting advertisers. On the other hand, you may have already bought another pair of socks and therefore the retargeted advert serves as more of a kick in the teeth than anything else. The key really lies in the proper implementation of sensible guidelines. Hopefully in 2011 that is exactly what we will see.