“We Look to Scientists to Settle Them”

A quick follow-up to my blog post on 4/4.  I just realized as I read this post by Laurent Bossavit, in which the author draws parallels between scientific method and the software development process discussion, applies to my question in my “Fool” post of how much we have to empirically evaluate our own opinions of “how to do.”  I’m not sure I was clear in that post, but I concluded that eventually we might have to “play the Fool” and just decide in the moment.

Reading Bossavit’s article increases my confidence in that conclusion as he says:

The trouble with opinions is that everyone has their own; you can always find one to suit any given prejudice. “Test-driven development reduces defect count”, says one expert; “test-driven development will wreck your architecture”, says the next.

Knowledge cannot be disseminated merely by everyone having a blog of their own. Blogs are great for voicing opinions – are they ever – and for having debates, but it’s unhealthy for debate to go on forever. We look to scientists for settling them.

Bossavit questions in this post our reliance on scientists to settle conflicting opinions, concluding that even scientists in their quest for empirically-supported truths are affected by factors that are not empirical whatsoever.  He teases out how language and opinion have a huge affect on what we accept as an “empirically supported” truth.

This post has been sitting in my queue of things to think about and write about.  What inspired me to finally publish it is today I saw some Twitter exchanges and found myself reading Dr. Alistair Cockburn‘s blog post called “Taylorism Strikes Software Development” in which his concern is through an empirical approach to work we can fall into the fallacy of a “Scientific Management” that characterized workflow management in circa 1910-1920. In the creative field we’ve thought a lot about this. I’ve seen images from the era of Taylorism used by folks like Made By Many in talking about why what we do today does not fit the manufacturing model of that time, and so I’ll reproduce one of those images here to visually make the point:

Model-T Ford Assembly Line

Model T - you should read about Taylorism if you are studying this for school. Model T expresses the old "Scientific Management" waterfall technique. New ways of working exist that are more fun! Look up "Agile Scrum " - scrum.org

Made By Many uses an image similar to this to illustrate the point that what we do as creative practitioners is not like factory production (see my own post here, their deck embedded). We’re not producing the same exact output time after time, so therefore we should not use a cookie-cutter process for production.

In “Fact and Folklore in Software Engineering”, Bossavit is addressing programmer productivity and looking at studies that aim to increase productivity by empirical examination of what works.  This view of empiricism is interesting to me as someone coming from the  marketing side, less from the development side, when it comes to making claims about ways to improve productivity.

Bossavit talks about making citations, linguistic modality, and how scientific research is itself affected by ideas about “what’s interesting” and other less-than-rational factors.

He mentions Bruno Latour, who wrote Aramis, and has a website called Mapping Controversies.  Two points Bossavit makes about Latour’s work were interesting to me:

Latour argues that the technology failed not because any particular actor killed it, but because the actors failed to sustain it through negotiation and adaptation to a changing social situation.

What Latour has highlighted is the pattern of progressive erosion, and eventual disappearance, of modalities as a statement becomes established as a fact in the sciences.

How I read these two statements is that, first, a method that included inspecting and adapting was not used in a given situation and so a project failed. In the second a “best way to think” or “best way to do” starts to become accepted through a particular use of language, rather than because it is 100% “true.”

Dr. Cockburn’s article addresses possible pressures on programmers to work in Lean or XP under a new “Best Way of Working” and compares this to Taylorism. He questions whether Lean and XP practices are a “better way to work” or whether we should say “everyone has to work this way.”

In the spirit of “Inspect and Adapt” (which are at the core of Agile practices, I do believe), it seems that “everyone has to work this way” flies in the face of a conscious view of what it is we do.

Ryder-Waite Hanged Man card

In fact, having been on a Tudors kick (at least for those episodes available on Netflix), I can’t help but be a little reminded of the Reformation and what seems to me to be the approach of “adopt our process or you’ll be burned at the stake.” (This is one reason why I wrote “I’m Not Gonna Force Agile on You.”)

We have to be careful we don’t end up burning people at the stake or hanging them because they don’t work according to our concept of “the best way to work.” The paradox is this might even include Agile approaches themselves. We might need to be prepared for cases in which we do not inspect and adapt, but rather follow a prescribed process for doing.  This may mean a level of awareness that even transcends “inspect and adapt.”  I guess it gets back to the old joke about what they say His Holiness the Dalai Lama said to the Hot Dog Vendor in Central Park (“Make me One with Everything.”)  And please don’t hang me for saying so. 🙂


Tags: , ,

3 Responses to ““We Look to Scientists to Settle Them””

  1. The Appearance of the “Last Responsible Moment” « Says:

    […] used the image of the Model T in a piece  called “We Look to Scientists to Settle Them, that talks about Laurent Bossavit’s article on the fundamental flaw of our reliance on […]

  2. @ericries , @zachnies talk @rallydev – live notes now – (reflections later) #agilenewlevel « Says:

    […] idea that we can be objective or empirical may itself be a flawed assumption.  See my post “We Look to Scientists to Settle Them” and “We Are Not As Smart as We Think” to see what I […]

  3. Reflections on the Webcast and Raw Notes: Conversation Between @ericries , @zachnies and @rallyon- @rallydev – #agilenewlevel « Says:

    […] idea that we can be objective or empirical may itself be a flawed assumption.  See my posts “We Look to Scientists to Settle Them” and “We Are Not As Smart as We Think” to see what I mean. We have so many […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: