The Google Dialogues: Search Neutrality

Sometimes I feel that the primary differences between Socratic Dialogues and Strawman Arguments are that: a) the former are apparently respectable, but the latter are fallacious, and... b) everyone since 399BC is not Socrates I too am not...


Sometimes I feel that the primary differences between Socratic Dialogues and Strawman Arguments are that:

a) the former are apparently respectable, but the latter are fallacious, and...

b) everyone since 399BC is not Socrates

I too am not Socrates - instead I am an unfair and imbalanced security geek who is strongly in favour of privacy, respects transparency and is vehemently against censorship, a juxtaposition which occasionally puts me into slight conflict with some of my likewise do-gooding peers.

A case in point is when I encounter organisations which are bent upon fixing Google, e.g.: a discussion coming up[1] at PICTFOR, the Parliamentary Internet Communications and Technology Forum, at the blog for which the following was posted:

Today, search engines like Google and Bing have become the Internet's gatekeepers, and the crucial role they play in steering traffic and revenues through the Global digital economy means they are now an essential component of the Internet's infrastructure. Should we look beyond network neutrality and consider "search neutrality": the principle that search engines should have no editorial policies other than that their results be comprehensive, impartial and based solely on relevance. Search neutrality has become an important topic because so much market power now lies in the hands of one company. Google's 90 percent share of both search and search advertising in the UK gives it overwhelming market power; both with search results and ad listings but some people think that Google, despite all the good services it provides, exploits it's gatekeeper status?

Something about that text must have irked someone, because shortly afterwards the posting was ripped out and replaced[2] with:

Foundem has an Antitrust complaint against google; it is essentially in two halves. The first concerns Google's increasing use of exclusionary automated penalties, which can remove sites from Google's natural search results. The second concerns Google's Universal Search mechanism, which transforms Google's ostensibly neutral search results into an immensely powerful marketing channel for its own services. Crucially, the placement of Google's own services is subject to different ranking algorithms and relevance signals than those used to place everyone else's. Because Google applies a different standard to its own services than to everyone else's, Google can favour it's own services. In preparation for September's US Senate Antitrust Subcommittee hearing, Foundem warned Senator Lee that Google would not admit to using different ranking algorithms and relevance signals to place its own services. Foundem have video for more on this:

...and perhaps unsurprisingly the text has again been edited and is currently[3] completely blank.

Now if Google are representing their search as ostensibly neutral then I very much would like to see their definition of neutrality, and perhaps have Advertising Standards look into that if there's a query; but I've never seen Google claim neutrality and frankly I would not want it.

Aside: I have seen them claim natural but that's an entirely different (and arguable) word.

I'm not the only one to critique the concept of search neutrality (PDF); but when I consider search neutrality what I hear in my head is a dialogue that goes a bit like this:

  • So what you're saying is that you can be a better Google than Google?

  • No, we're saying that Google are wrong.

  • Because of how they order their search results?

  • Yes - it's wrong. They shouldn't be allowed to order search results any way that they choose, thereby promoting their agenda.

  • You say that there's a natural order to Google search-engine output; but I've written small search-engines and attest that there is nothing natural about the output of PageRank or whatever they're using to rank results this month; plus I remember the development of AltaVista search, the output of which eventually sucked. Result ranking is a feat of engineering subject to infinite tweakage. At what point should that be stopped because it's unnatural?

  • When it's for Google's gain, and becomes anti-competitive!

  • So why not create your own search engine that is better and promote that? Or use Bing? Or support Duck-Duck-Go which is tremendously ethical?

  • Because Google are here, they're wrong and they need to be fixed.

  • Yes, you said that. So Google should listen to you regards how to order their search results?

  • Yes.

  • Why?

  • Because their ordering is not fair and it promotes their own content, such as Google+!

  • And... ?

  • ...and we have to use Google because it's a monopoly!

  • But it's not really a monopoly. If people want fuel or food and were Google the only source of that, then it would be a monopoly. Were webspace to be a finite resource and Google a gatekeeper to having any web presence at all, then it would be a monopoly. But there are other search engines, and you can keep adding more, and people can use alternatives for everything that...

  • That's irrelevant. Google have massive market share. We need regulation so that Google orders search results fairly.

  • You mean: in the way that you want them ordered?

  • No! The way that we need them ordered! A fair way!

  • But who is we? I want search results to be ordered the way that Google wants to order them. Google are pretty smart and in a position to change their systems if they don't work out. Anyway - if a search doesn't work out for me then I'll ask a friend for help, or ask Twitter, Facebook, or bounce around a few blogs. I'll go elsewhere, and if Google don't like that then they will fix the issue or lose my custom. This is how markets work.

  • Ah, but you're an elitist!

  • I'm a what?

  • You understand how to do that - how to find alternatives. The normal users of the web don't know about alternatives, and so they are disadvantaged...

  • Wait, so... a) you should be put in charge of Google's search ordering because there are people out there who are new to the net, and b) doing so would be OK because it will ever be thus? Have you actually met any teenagers recently?

  • Well it won't be me in charge of course, there will be a committee...

  • You're avoiding the question.


  • Actually it does, I just checked. Third result. After her home pages. If ever it did not then perhaps they fixed that, like I said. Google will respond better to public shaming and demand than regulation.

  • What we need is some sort of oversight committee who can tell Google how best to order search queries and what to leave out!

  • Whoa - now you're talking about censorship?

  • No no no... no of course not, this is all for good! For the benefit of everyone!

  • But you just said what to leave out?

  • Yes - after all, some content self-evidently should not be promoted through search results - piracy, obscene material...

  • But you don't want Google to be making that call about their own service?

  • Of course not! They're not fair, and they won't know what should be elided because we haven't told them yet!

  • We already covered that - and you still haven't justified that your so-called problem will still be an issue in ten years time, let alone five...

  • You realise of course that you're trapped inside Google?

  • In a sense, but only in as much that I choose to use some of the services they offer; not all of them, but that's fine. There are alternatives.

  • You're a slave to an invidious info-chthulhu that consumes all your data and profits from it!

  • Now you're just being dramatic - I make informed decisions about where to draw the line with Google-use; maybe we should inform and educate other people rather than trying to regulate a dynamic technology sector? Stuff like how to be reasonably anonymous on the net, when you need? or How to monetise your own data?

  • That's not possible.

  • And what happens when Google tweak their ranking algorithms - does your committee have to approve the changes to the code?

...and on it goes. James Grimmelmann's paper (linked above) takes these arguments to a far more nuanced level, but they suffice for me.

It should be obvious that the response you get to any question will be coloured by whom you ask, in much the same way that I wouldn't ask vegan friends for pork recipes; but that doesn't mean I should seek to normalise all my friends to all provide the same responses.

To do so would be boring - and hideously easy to manipulate.

Follow me as @alecmuffett on Twitter and this blog via the RSS feed.

[1] 6th Feb at 1830h, The Grand Committee Room, Westminster Hall
[2] NetNewsWire has a feature to highlight edits to postings. Very useful.
[3] as of 19th Jan at 2315h