Misattribution
It’s time we spoke of misattribution. We have discussed the meanings of many things and the constellations of meaning in a given situation. Misattribution is the attribution of a meaning that is simply wrong.[1]
Supermarket theory
With a hat-tip to Patrick Hoverstadt, a good example of misattribution comes from the world of supermarkets. What do you think of the idea of giving supermarket branch managers a bonus based on the sales or profits achieved by their branch? Well, no prizes for guessing that we wouldn’t be writing this blog post if it was a good idea.
Let’s step back to understand the meaning of this idea. It rests on the notion that a good branch manager will achieve higher sales or profits than a poor branch manager. So, we need to look at the ways a branch manager might affect sales or profits. The layout of the store and what effect the layout has on buying behaviour is definitely done by head-office types. The selection of produce to be sold and the price that it is sold at is also definitely a head-office function. Even more obviously, the siting of the store and its convenience (or not) for customers is not theirs. We could go on.
If branch managers do not wield the major tools that are thought by their head-offices to have the major effect on sales and profits, then it doesn’t make sense to use the branch’s bottom line to design incentives.[2] But of course people do, because it is obvious, isn’t it, that the bottom line is the thing!
Most large companies reckon to drive their decision-making by hard data.[3] Clearly there is a problem if the data does not mean what they think it means. This is true even when they’ve made an effort to ‘conform’ the data. When people with sophisticated statistical understanding have studied this systematically it turns out that maybe 80% of the meanings ascribed to data are misattributed.
80% is a handy sort of number, think of the 80/20 rule.[4] I have also seen a recent study that said that in a large scale evaluation of theories about diet, all based on epidemiology and association between population health and diet, 80% of the theories turned out to be wrong. They were much more likely to be wrong than right.[5] When someone tells you how much fish they eat in Japan, just pretend to be interested for a while. If you get irritated, just ask them about association and causation. This transfers the irritation.
Manipulation
There is unconscious misattribution and there is conscious creation of a misattribution narrative. If we take the history of big tobacco, of big oil, of big pharma, of big food[6] we can see the techniques being refined and more generally applied. It is far more cost effective[7] to establish the false narrative and get people to spread it ‘for the good of the world’ than it is to simply promote your products against a background of information casting doubt on whether they should be sold at all.
Once a suitably misattributive narrative is out there, then people can pile in to support it in the name of research. We explored previously the “cows cause global warming” narrative which is completely devoid of either science or common sense but which has caught on massively around the world with people who anthropomorphise the animals that humans eat.[8] We also explored projection and the ways in which emotional gotchas, like that anthropomorphism, open us to being manipulated.
We have also explored this using the terminology of memes and their propagation. The design of memes that will multiply and cascade around the world is probably now the core skill of people trying to make unacceptable industry behaviour acceptable enough. There is a demonstrable cross-fertilisation and inter-working of people from these industry sectors so that we see memes that have a pull effective in various dimensions. Cows causing global warming diverts attention from big oil, it distorts people’s diet into snacks big food wants to sell them, causes illnesses that big pharma want to be prevalent… in short, everybody wins.[9]
These seven traps describe the ethical hazards in all artificial intelligence.[10] But they can be applied to any development of memes. It is precisely the appeal to technical “facts” and “obviousness” that is used to disguise the political intent. The memes have to have a general appeal, which we know from marketing theory means they have to be as bland as possible.
- the reductionism trap
- the simplicity trap
- the relativism trap
- the value alignment trap
- the dichotomy trap
- the myopia trap, and
- the rule of law trap
Influencing theory
Some people trace modern influencing back to a Nike advert in the 70s. It featured a great photo of a runner in silhouette by a lake. The caption — “the trail has no end” — was hugely successful in selling trainers, despite the fact that they couldn’t be seen in the ad and the caption/slogan does not have any obvious meaning. People found the meanings in the ad that they needed to find, without Nike ever making any claims.
Steven McKevitt, Visiting Professor of Brand Communication at Leeds Business School, speaking on Thinking Allowed on BBC Radio 4 said that, despite his lifetime academic interest in and study of influencing, he knew that he was completely saturated in messages that he was not any better able to resist than anyone else. We are totally immersed in professionally propagated influencing, and they are not at all limited to commercial goals: political campaigns for instance are possibly more prevalent, and many political influencing objective are socially divisive to allow political projects to reach their goals.
Aidan can remember talking to his father-in-law about this when he was still in his twenties. Was it possible to read the Daily Telegraph — for the sports reporting — without becoming imbued with Daily Telegraph political values[11] and not be aware of how you were being influenced. I grew up without a television, and have lived my whole life without one, because of my sense of this process. I cancelled my embryonic Facebook account because of the evil being spread by Facebook, but I had the same sense of being caught up in something that was not possible to bring under my conscious control.[12]
As I walked around Sainsbury’s this morning I was especially troubled, writing this blog, by the rampant lies, misinformation, manipulation, influencing that is intrinsic to the marketing of food products. Can supermarket shoppers really process out the notion that low fat is somehow high value? That breakfast cereals are what everyone eats for breakfast? Even some of my otherwise sceptical friends proudly declare their breakfast of cereal and low fat milk as being healthy. This diet is known to be the best way to fatten pigs.
Why does this sea of influencing leave us drifting on whatever currents come our way? Well, precisely because we misattribute the motivation of the influencers. We say: I know what they are trying to make me do and I am not that stupid or that blind. What it means for influencers to be two or three steps ahead is precisely that they have thought through what our reactions will be and have designed their strategies around those predictable reactions.
Let’s go back to the influencing being political, whether it is overtly political or not. The sort of people who get to design and approve influencing strategy are not your friend. And they are not trying to improve your life or the society you live in. They are out to exploit what little you have. Period. Our colleague Angus Jenkinson tried for a year or more to convince Aidan that it was possible for corporations to be honest, beneficial social citizens. That it was just a matter of getting them to enact values that they actually cared about for everyone’s good. But I think in the end he was still about convincing customers and the public about the beneficial effects of what a company was doing. Influencing to the end.
We usually get to a meta word and the meta-influencing is precisely what Angus was attempting, and he is a persuasive man. Meta-influencing says: it is possible for influencing to be used well and responsibly. I can only say: show me. Influencing can only be about me knowing what is best for you.[13]
Reclaiming ourselves
My colleague Michael Jacobs says that many people use self-touching to establish their existence to themselves.[14] We swim in a sea designed to prise loose our hold on ourselves. Just as medicine needs us all to be ill, influencing needs us all to be lost. That is the real meta story.
There is a Canadian professor of the social aspects of technology called Albert Borgmann. He says that we should understand technology from the perspective of human focal practices. A focal practice is precisely something that grounds us in our ancient humanity and so provides a way of understanding how technology changes our world.
The paradigmatic example is the log fire providing warmth. While it is perfectly possible to have underfloor central heating that is far more convenient and controllable than an open fire, the fire has many more connections to our social existence. People have always had a connection to the natural world gathering firewood. They have always had exercise in cold weather chopping wood. And they have always gathered around a fire to talk into the night.
Another example he uses is wilderness. The whole point of wilderness is that it is wild and untouched by human development. It cannot be improved because any intervention to make it more accessible or more comprehensible or safer or more controllable can only detract from its wildness.
We stay in contact with our practical humanity by using the focal practice perspective. We can easily see that any and all influencing can only take us away from ourselves and our real relationships the way interventions detract from wilderness. They are not us, and while they may be attractive and even useful, that comes at the price of losing clarity about who we are.
As we have commented before, the result of all the recent political influencing campaigns, even without their blatant illegality, it to drive a wedge between people, to destroy relationships and trust. Classically, to divide families. If we think back to that Nike advert, this is implicit. If we all find ourselves and our projections confirmed in these campaigns, we strengthen our false consciousness at the expense of our empathy with others. Whether we find that Brexit calls us to a new sense of national purpose or underlines our instinct that destructive political mischief is afoot, we are not engaged in a political process that can find ways to deal with the real issues such as migration.
Focal practices are not simply a philosophical perspective, useful though they are as such. Sitting around a log fire talking into the night is not simply a thought experiment, and if it remains one will lose the focal practice itself. No, we need to actually sit round that fire together, actually get our hands dirty in real soil, actually lose ourselves physically in the wilderness. And we need to pay really close attention to what happens when we do. We must pay attention to our sense of ourselves and the world, to our relationships, to our perspective on those sweeping, powerful campaigns to turn us into something else. The risk of losing ourselves is absolutely real and present. The acute symptoms of that are things like school shootings and opioid deaths sweeping the US.
If that juxtaposition of sitting round a fire and shooting your schoolmates feels strange, then you have the beginnings of the motivation you need to find your bearings.
When I speak to people about my total abstinence from television, they invariably say: yes but there is this wonderful programme that I wouldn’t want to miss. Which is like talking about the utter convenience of underfloor central heating and your ability to switch it on from the car on the way home. But that is to miss the importance and centrality of certain other things like the fire to sit around. The ultimate antidote to influencing is to be centred somewhere that makes it obvious that influencing is always and everywhere an intrusion.
— — —
[1] Misattribution in the sense of something fundamentally wrong, as opposed to the attribution of a mostly right reason to the wrong party…
[2] Some of those decisions might be deemed too important to be left to the local people, or require analytical tooling, or simply be imposed from the centre for reasons political.
[3] And data is indeed hard. We could write for days (have written for days?) on the trouble caused by data-driven decisions and evidence-based policy. The world is not as flat as the data suppose. For an alternate but compatible view on non-linear data analysis, you could read Taleb. (Less gentle, equally serious?)
[4] The 80/20 rule itself is wildly misunderstood, perhaps because the two numbers happen to add up to 100 but are percentages of different things: 80% of something attributed to 20% of something else. In a particular situation it could equally be 80/5 or 95/20…
[5] Which doesn’t make them useless. Arguably the proper application of a scientific method relies upon a lot of wrong things and a few continually-challenged probably-right things.
[6] Do you see a theme? It might be irresponsible to insist that small is beautiful, but certainly too-big-to-fail is a tragedy in the making.
[7] It’s also more cost effective to pay staggeringly large fines. Staggeringly large from the perspective of an individual, but insignificant from the perspective of a global giant. This may be changing, given regulations like GDPR that focus on percentages of global revenue. Needs a test case or two to determine effectiveness.
[8] As perhaps we should, as far as them being fellow sentient beings and all that. But perhaps no further than thanking them for providing an excellent meal. Because there are differences — from a recent post, herd animals huddle together for comfort and safety, whereas predators like us feel anxious when crowded. It’s hard enough appreciating differences between human viewpoints without trying to cross species… Also, my friend Tom, formerly chief exec of the Rare Breeds Survival Trust, urges us to routinely eat the critters that we want to survive in the world.
[9] Ok, not everybody.
[10] Interestingly, Google have recently both convened and disbanded an ethical committee for AI. It seems that their employees had objections to the political affiliations of some of the members, regardless of their technical experience for the matter at hand. This particular list of traps comes to us from https://freedom-to-tinker.com/2019/03/25/ai-ethics-seven-traps/
[11] The Daily Telegraph, of course, famously, is read by people who think the country is being run by another country.
[12] Contrariwise, the other “I” of this blog also grew up without a television, and then with one, and now live blissfully without. My facebook account still exists, but has been unused since at least May 2018, because every time I think to use it I’m encouraged to accept their GDPR-updated terms and conditions.
[13] The same, of course, is true of laws and rules and other more overt forms of influence. You will keep your dog on a lead because you’re not to be trusted to judge whether your dog should be on a lead. By creating laws we create criminals, and all that.
[14] Self harm being a less benign form, I suppose.
