Proposed to IETF: Use 'Deception Patterns' instead of 'Dark Patterns'

In a recent Fediverse discussion about Dark Patterns in Games I was pointed to the non-inclusive terminology use of “dark” in Dark Patterns by fellow fedizen El Joa. Now “Dark Patterns” is terminology that is very often used on the internet. I thought of an alternative and suggested we could use Deception Patterns instead.

This morning I encountered more uses of ‘dark pattern’, this time as hashtags, so I decided to write a proposal to the authors of IETF’s “Terminology, Power and Oppressive Language” Internet-Draft, Mallory Knodel and Niels ten Oever, to consider adding ‘Deception Patterns’ in a future version of the document.

Below you can see the gist of the email I wrote:

First of all I want to thank you for publishing the IETF Internet Draft “Terminology, Power and Oppressive Language” ( https://tools.ietf.org/id/draft-knodel-terminology-00.html ). It is a resource that I spread very often, and I gladly use the suggested terminology myself.

Secondly I want to mention that - in light of recent uptick of attention to using inclusive language in society - I hope that soon an updated version will be published. I find that reference to an official recommendation works best in having others embrace the alternatives, and make them the new normal.

It is delightful to see that code forges such as Gitea ( https://gitea.io ) were early adopters of using ‘main’ as the default branch, instead of ‘master’, and others - such as Github - intend to follow. Another example of improvement is the switch of many Ad-Blocker projects to ‘blocklist / allowlist’ terminology.
​​
Now for the reason of this mail: Dark Patterns ( https://darkpatterns.org )

Though the referenced website refers to them as “tricks used in websites and apps”, with the growing tech awareness their connotation aligns more to “the evil and deceitful ways that surveillance capitalism enterprises use to hijack your attention and get you hooked to their products and services”. In other words a clear ‘Dark / black equals Deceit / Evil’ relationship.

Clearly this terminology is up for a change as well.

I hereby propose to use the alternative of Deception Patterns instead.

Why deception? Well, the behavioral scientists and UX designers who craft them have the clear objective to mislead the user base in following behavior patterns they did not intent to follow. And throwing all kinds of human psychology tricks into the mix to do so.

So there you have it. Yet one more small step towards inclusivity in language use :slight_smile:
I hope you will consider this in future versions of the IETF draft.

If you have other examples of non-inclusive language in technology design, then I encourage you to create a topic in the Inclusion category to make us all aware, and we can discuss alternatives.

Thank you.

#HumaneTechNow
(Credits: Photo by Sharon McCutcheon from Pexels)

3 Likes

I received a response from IETF member Mallory Knodel. IETF has a newer version of the terminology document and a Github repo to PR to. So I just did that for the Dark Patterns alternative.

I informed Fediverse, Twitter and our HTC LinkedIn Group of the new terminology. Furthermore the great founders of Greaterthanlearning - @m3me and Nate - have committed to using Deception Pattern from now on.

Will you help promote Deception Pattern as the new norm? Just do it!

2 Likes

Hey @aschrijver, awesome initiative, thanks.
As a translator and PhD student in area of credibility evaluation,
I’m all for conscious language shaping in the direction you suggested! kudos.

To make the language more specific, I would not go for ‘deceptive’ right away, since some of the patterns described are not pretending to be something else (like an impostor website), but making things deliberately hard for users (although some labels were deceptive or misleading).
So, maybe a term along the lines of obnoxious, exploitative, unfair, obstructive, imprisoning, hassle, harassing elements …
Don’t know what the best option will be and this might warrant some discussion.
My four considerations for the term choice would be:

  1. specific enough to describe the wrongdoing (and distinguish it from others)
  2. purpose/function: How will the term impact the business (i.e., if shaming or warning is the desired outcome, is it strong enough to convey that)?
  3. ease of use.
  4. coverage: all desired types of ‘dark’ patterns included.
    (Also notice how the criteria form the mnemonic SPEC. ^^)
    ‘Deceptive’ for me checks some of the boxes: 1. n, 2. y, 3. y, 4. y/n.
    All my suggestions above check some other boxes, but not all.
    Currently, I like ‘harassing’ patterns best (1 y/n 2 y 3 y 1y/n).
    That’s the outcome of continued unwanted interaction feels like. Terms from the area of (unwanted) relationships could work well. The forced face recognition example would be (data) ‘stalking’. It feels like there is a comprehensive term just around the corner, once seen the complete list…

Cheers, Dimitri

2 Likes

This was a great add, @aschrijver :ok_hand:
“Deception Patterns” defines the practice very well.

I’ve had my attention called while using the non-inclusive term once.

1 Like

Thank you @ibaldo.

@Dimitri-Molerov I see what you mean, and agree to your observation on the coverage in those four points.

But not the alternative you suggest of ‘harassing’. The vast majority of these patterns are carefully designed such that you do not notice them, while they steer you in a certain way that the designers want to steer you to. And that is usually a place where user needs are less well served (e.g. giving up privacy, mixing in an Ad, etc.). Harassing implies the opposite of that. When you are harassed, the pattern would be right in your face, adding friction to the UX.

Note that there is a difference too, between ‘deceptive’ and ‘deception’, as the first can also be unintentional, but the latter conveys intent.

The problem with finding the good alternative, is that ‘dark’ constitutes such a broad umbrella term. In order to find a good alternative a word with equal broad coverage would have to be found, or something that comes closest. The alternative of having multiple alternatives to denote different categories of ‘dark’ pattern would also not be optimal.

1 Like

perhaps covertly manipulative?
or just manipulative?
coercive (but also implies intent to some degree)
disingenuous is always a great word and seems to me to fit but then I’m a newb

1 Like

Manipulation patterns could be a good second alternative you could suggest to the repository. That is I think a name that is very descriptive and maybe even covers things more broadly than deception patterns.

There’s a pull request to the IETF Terminology repository to fall back on the NIST guidelines for inclusive language: IETF | IESG Statement On Inclusive Language

Guidance for NIST Staff on Using Inclusive Language in Documentary Standards