Twitter's founders launch two new services. What are they, and do they have a hope?

Medium and Branch could be third (and fourth) time lucky for Stone and Williams.

Ev Williams and Biz Stone, the co-founders of Blogger (now owned by Google) and Twitter, have launched not one, but two follow-up projects, Medium and Branch.

The two men will be staying on as directors of Twitter, which poses a problem for them commercially – how do they use their expertise to carry on the string of hits, without cannibalising their previous business? One of the Twitter's cofounders, Jack Dorsey, decided to abandon the social media sector entirely, instead attacking two monopolies at once with Square, a platform which allows anyone to accept credit card payments with just an iPhone or iPad.

And with their audacious move to launch two start-ups at the same time, Stone and Williams have that problem doubled. Will people really use Twitter, Medium and Branch at the same time? That's the plan.

What are they?

First things first: what exactly are the new platforms? It's always going to be tricky to describe these things until usage patterns have built up around them organically. You inevitably find yourself resorting to analogies with existing services, which can be far from perfect: I remember, in 2007, attempting to describe Twitter to someone as "like Facebook status updates but without the rest of Facebook". Needless to say, I did not convince them to sign up.

Stone and Williams seem to have a firmer idea of what a mature Branch and Medium will look like than they did with Twitter, however. The latter famously was heavily driven by its users, with conventions like hashtags, retweets and @-mentions invented on-the-fly, and then incorporated into the architecture of the site later on. The way people use it today bears little resemblance to the way they did five years ago.

Medium is a very image-centric platform for content grouped around specific themes. The idea is that users create certain "collections", which are grouped around a theme. Sometimes, these collections are closed, but they can be open to extra contributions. Williams explains (on Medium, of course):

Collections give people context and structure to publish their own stories, photos, and ideas. By default, the highest-rated posts show up at the top, helping people get the most out of their time in this world of infinite information.

Together, the contributions of many add up to create compelling and useful experiences. You may be inspired to post one time or several times a day—either way is okay. If you’re more ambitious, you might create a collection of your own.

Collections exist on topics like editorials, things people have made, nostalgic photos and crazy stories, while the site has a voting function which, ideally, ensures that interesting contributions to those collections float to the top.

Although the design is focused around images, and reminiscent of Pinterest in its gridded layout, posts can be all text, and can indeed be quite weighty. In terms of the (small-m) medium, Medium looks to be encouraging a similar approach to Tumblr (although with much more high-brow content, ideally). Lots of images, some text, and a few links out. The idea is that the individual posts become something more when the group as a whole takes over.

Branch is far more about the conversation as a whole. At its heart lies a question and answer format similar to Quora, another Silicon Valley darling. Users start conversations with an opening post, and can then invite others to join in. The chats are readable by anyone, but only invited users can contribute - but, importantly, anyone can click on any post to "branch" it into its own thread.

Topics being discussed at the moment include today's changes to Twitter's platform, TEDx, an offshoot from the popular TED conferences, and Obama's re-election prospects.

It's easier to describe than Medium, but that's partially because it's a far simpler service. It knows what it wants to be, but there's far less chance for users to discover.

How do they work with Twitter?

If it wasn't clear before that these sites need to work with Twitter, rather than against it, the company today announced changes to the way they deal with third-party apps and services which appear to be a precursor to banning many of them from the network entirely.

Branch is most explicit about how it would mesh with Twitter. It sees itself as a way to take those long, unwieldy five- or six-participant conversations off-site to somewhere where arguments can be developed in a bit more length. As seen in this discussion, it even encourages you to embed tweets to begin the chat.

Medium targets itself at a different sector. It still links to Twitter - right now, the only way to sign up for an account is to use your Twitter account, for instance - but there are few explicit connections between the two services. Its target is different, lying somewhere between Tumblr and Pinterest. The most interesting claim the founders make about it is that it will not require massive engagement to get noticed on - which is a problem with both those sites. If everything works as stated, then a first post could become the most "interesting" one on the most-read board. In this, as with its voting mechanic, Matter actually bears more than a passing resemblance to Reddit. Submit cool things, get up-votes, and be read by the crowd, all of which is fragmented over boards which anyone can create.

Reddit, of course, co-exists admirably with Twitter, so there should be no problem there.

When I wrote on Twitter's API changes, I argued that even worse than the ill-thought out rules being strictly applied is if they aren't strictly applied – if, as there are indications, Twitter gives "good" sites an easier ride.

Sadly, Branch just adds to that notion. While the site will doubtless play well with Twitter, it breaks several of the company's design guidelines (soon to become design requirements). Tweets are displayed without retweet, reply, or favourite buttons, names are displayed without the username next to them, and the Twitter logo is not always displayed in the top right corner. Despite this, something tells me it will not have its API access revoked.

Do they have a hope?

The real question, of course, is whether these things can grow beyond the initial hype. Are they filling niches that need to be filled? Can they encourage users to switch from competing services? And will they work as they scale?

Of the two, Branch is the one which has the more obvious chance of success. It is easy to imagine people saying "let's take this to Branch" when a conversation on Twitter gets out of hand, and the integration between the two services makes that something even the least technologically-minded user can do. Obviously the "featured branches" view of the site would gradually fade into the background as it grew, just as you can't get a whole site feed for Twitter anymore, but this is to be expected; as Dalton Cadwell argued, the global feed is useful for avoiding anti-network effects (where a site gets less useful the more people are on it; compare, for example, Yahoo! Answers and Quora) in a growing site, but useless once something reaches critical mass.

Medium is a different beast entirely. Its problem is getting people to use it. Is it a Tumblr replacement? Pinterest? How should you get content into, and out, of it? Is it for ephemeral posts, or will it have a working archive?

Yet if it does work out – if people do start sharing wonderful things, and telling each other "nice work!" (the equivalent of an up-vote, to use the Reddit analogy) – then Medium has a chance of being, not just a useful addendum to other social networks, but a hub in its own right. Reddit has 35 million users, and an incredibly engaged community. Who wouldn't want a piece of that?

 

Medium and Branch.

Alex Hern is a technology reporter for the Guardian. He was formerly staff writer at the New Statesman. You should follow Alex on Twitter.

Alan Schulz
Show Hide image

An Amazonian tribe is challenging scientific assumptions about our musical preferences

The Tsimane’ – a population of people in a rural village in Bolivia – are overturning scientists' understanding of why humans prefer consonant sounds over dissonant ones.

It was 29 May 1913. Hoards of Parisians packed out the newly-opened Théâtre des Champs-Élysées. Messrs Proust, Picasso and Debussy were in attendance. Billed for the evening was the premiere of Le Sacre du PrintempsThe Rite of Spring, a ballet and orchestral work debuted by Russian composer Igor Stravinsky.

The attention and conjecture focused on the theatre that day meant expectations were high. However, within moments of the piece beginning, all preconceived notions held by the audience were shattered, as what was unfolding in front of them was a musical tragedy unlike anything they had ever witnessed.

A bassoon hummed into the ether before ballet dancers stomped on stage; the music, unpredictable with its experimental edge, drove forth the onstage narrative of a young girl whose selection during a pagan ritual saw her sacrificially dance towards death. Stravinsky’s composition and the ensemble of the night caused the room to descend from laughter and disruption to chaos and uproar.

The employment of dissonance – sharp, unstable chords – largely contributed to the audience’s disturbed reaction. Dissonant chords create a tension, one which seeks to be resolved by transitioning to a consonant chord – for example an octave or perfect fifth. These musical intervals sound far calmer than the chords which riveted the audience of The Rite of Spring.

Dissonant and consonant intervals find themselves as binary opposites; the frequencies at which notes played together vibrate determine whether an interval is consonant or dissonant. Consonant intervals have simple mathematical relationships between them, but greater digression from that simplicity makes an interval increasingly dissonant.

It’s long been believed  both experimentally and anecdotally – that the preference among Westerners for consonant chords highlights a universal, perhaps biologically-rooted, leaning among all humans towards consonant sounds. If you were present at the introduction of Stravinsky’s The Rite of Spring on that night of furore in Paris, you’d find it hard to disagree.

There is, however, a growing movement against this consensus. Ethnomusicologists and composers alike argue that favouring consonance may just be a phenomenon that has evolved from Western musical culture. And following the visit of a group of researchers to a remote Amazonian society, these claims could well be grounded in scientific evidence.

Led by Josh McDermott, an MIT researcher who studies how people hear, the group travelled to a village in the Amazon rainforest called Santa Maria. It’s populated by the Tsimane’ – a group of native Amazonians whose rural abode is inaccessible by road and foot, and can be reached only by canoe. There are no televisions in Santa Maria and its inhabitants have little access to radio, meaning exposure to Western cultural influences is minimal.

The researchers were curious to see how the Tsimane’ would respond to music, in order to determine whether they too had a preference for consonant sounds over dissonant ones. To everyone’s surprise, the Tsimane’ showed no preference for consonance; the two different sounds, to the Tsimane’ at least, were equally pleasant.

Detailing their research in a paper published by Nature, the group explains how the Tsimane’ people’s indifference to dissonance is a product of their distance from Western culture and music, removing any purported notion that humans are hard-wired to praise perfect fifths and fourths.

McDermott tells me that the Western preference for consonance may just be based on familiarity. “The music we hear typically has more consonant chords than dissonant chords, and we may like what we are most exposed to,” he says. “Another possibility is that we are conditioned by all the instances in which we hear consonant and dissonant chords when something good or bad is happening, for example in films and on TV. Music is so ubiquitous in modern entertainment that I think this could be a huge effect. But it could also be mere exposure.”

To fully gauge the Tsimane’ responses to the music, 64 participants, listening via headphones, were asked to rate the pleasantness of chords composed of synthetic tones, and chords composed of recorded notes sung by a vocalist. At a later date, another 50 took part in the experiment. They had their responses compared to Bolivian residents in a town called San Borja, the capital city La Paz, and residents in the United States – locations selected based on their varying exposures to Western music.

What made the Tsimane’ particularly interesting to McDermott and his group was the absence of harmony, polyphony and group performances in their music. It was something the researchers initially thought may prevent an aesthetic response from forming, but the worry was quickly diminished given the Tsimane’ participants’ measure of pleasantness on the four-point scale they were provided.

Unsurprisingly, the US residents showed a strong preference for consonance – an expected preference given the overrunning of Western music with consonant chords. Meanwhile, the San Borja and La Paz residents demonstrated inclinations towards consonant sounds similar to the US residents. The implication of these results – that consonance preferences are absent in cultures “sufficiently isolated” from Western music – are huge. We most probably aren’t as polarised by consonance and dissonance as we assume; cultural prevalence is far more likely to have shaped the consonant-dominant sounds of Western music.

McDermott raised the question about why Western music may feature certain intervals over others to begin with:

“One possibility is that biology and physics conspire to make conventionally consonant and dissonant chords easy to distinguish, and so that distinction becomes a natural one on which to set up an aesthetic contrast even if the preference is not obligatory. We have a little evidence for this in that the Tsimane' could discriminate harmonic from inharmonic frequencies, which we believe form the basis of the Western consonance/dissonance distinction, even though they did not prefer harmonic to inharmonic frequencies.”

There has been some criticism of this. Speaking to The Atlantic, Daniel Bowling from the University of Vienna said:

“The claim that the human perception of tonal beauty is free from biological constraint on the basis of a lack of full-blown Western consonance preferences in one Amazonian tribe is misleading.”

Though the results from the Amazonian tribe demonstrate a complete refutation of previous assumptions, people's musical preferences from other cultures and places will need to be analysed to cement the idea.

With research beginning to expand beyond WEIRD people – those from a Western, Educated, Industrialised, Rich and Democratic background – the tastes in music of people the world over may continue to surprise, just as the Tsimane’ did.

The Rite of Spring, which was met with ridiculing reviews has now been canonised and is considered to be one of the most important pieces of music of the twentieth century. A Tsimane’ crowd on that tender night a century ago in Paris may have responded with instant praise and elation. With further research, the imagined Bolivian adoration of a Russian composer’s piece in the French city of love may prove music to be the universal language after all.