Blog: Lightning Talk on Emotions and Oracles

I gave a lightning talk at STAR East on emotions and oracles. Oracles are the principles or mechanisms by which we recognize a problem. Jerry Weinberg has written a lot about the significance of emotions in our work, but I’m aware of few other people who address the matter so directly in the software business.

Over the last little while, I’ve begun to find that there’s an emotional component to finding bugs, and this talk is the result.

If you’ve got PowerPoint, download the PowerPoint version, load it up, and hit Shift-F5; it’ll clip along at a brisk pace automatically. If you don’t have PowerPoint, download this bundled presentation instead, and it’ll run automatically too.

Want to know more? Learn about upcoming Rapid Software Testing classes here.

8 responses to “Lightning Talk on Emotions and Oracles”

  1. JonathanKohl says:

    This is great work Michael.

    “Brilliant!” 🙂


  2. Shrini Kulkarni says:

    >>>>Why do so many in our profession seem to be so oblivious to the value of emotions?

    Why? Because emotions cannot be

    – Quantified
    – Measured
    – Improved
    – Trivialized into a 90-minute class room session
    – Institutionalized
    – Sold as best practices
    – Standardized
    – Repeated
    – Scripted

    Emotions can’t be easily reconciled with

    – Industry standards
    – Metrics
    – Productivity Data
    – Historical data
    – Certifications
    – Process maturity models
    – Some basis for comparing two testers
    – Tools
    – Regression models


  3. Jon Bach says:


    One of the Top Ten Tendencies I’ve seen in testers over 12 years of hosting interviews is Dismissed Confusion. They recongize a problem, but figure it must be them because there’s an emotional component to the discvoery, which must not be allowed or actionable, right? After all, this is testing, and we all know there’s no crying in Testing! Any sign of weakness is a bad thing to demonstrate in the interview, and often, emotions are considered weak.

    A little bit of self-criticism is healthy, but too much is a pathology.

    I get the feeling that many test managers and executives out there are fueling this “No Crying in Testing” mentality by staffing programmers in Test with the aim that the programmer just needs to think one time and get it overwith! Then… “Program that thinking stuff into the auotmation suite so that none of that loosey-goosey humanity stuff need happen again! After all, there’s no crying in testing, we have software to ship!”

    We need a lot more of what you’re saying with this post.

    Excellent work, Michael.

  4. Michael says:

    Thanks, all, for the kind words and especially to Jonathan Bach and Shrini for their amplifications of the ideas here. Also thanks to Jonathan Kohl for the related stuff on his blog, and for provoking me to post the talk here; Jonathan has a knack for amplifying on my work before I publish it, which is very stimulating.

    Thanks in part to the reading that I’ve been doing lately (How Doctors Think, Stumbling on Happiness, Mind Hacks and Mind Performance Hacks), I believe that there’s lots more work to be done in this area. Cem has also encouraged me to dip my toe in the water of psychophysics, and one day I’ll dive the rest of the way in.

    There’s this wonderful quote from Goethe about learning to see with a feeling eye, and feel with a seeing hand; I think that may be a good metaphor for a great tester’s aspirations.

  5. Jason says:

    I’m not sure why people think that emotions cannot be quantified or measured.
    The video games industry certainly designs their games to have differing emotional effects are differnt stages of the games. They have developed ways of representing emotions and their “quantity”. is a paper on the “emotional requirements of video games” By David Callele (sorry, but it’s an ieee paper) That has a easy to use system for representing desired emotions, that could be used to represent the emotional state a test during/after a session of testing.

  6. Michael says:

    Hi, Jason… thanks for the feedback.

    I’m not sure why people think that emotions cannot be quantified or measured.

    One answer is that emotions are intangible. You can’t touch ’em, and you can’t put ’em in a box. There’s a difficult measurement problem with things that don’t have physical attributes to measure. For emotions, there are some physical stand-ins that we can use–blood pressure, galvanometers, EEGs, MRIs–but the experience of an emotion is subjective–and I believe more immediate than the timescale measured by these devices (with the possible exception of EEGs). Check out Stumbling on Happiness by Daniel Gilbert for an accessible discussion of the issues. We can also use opinions (or surveys of opinions), but I’d suggest that the more useful form of information in those kinds of study is assessment, rather than measurement.

    The video games industry certainly designs their games to have differing emotional effects are differnt stages of the games. They have developed ways of representing emotions and their “quantity”.

    Note that you yourself put “quantity” in quotes just there. Yes; you can model emotions, but quantitative measurement systems for them seem unconvincing to me.

    I’d be interested to find out more about how the video companies do this, other than by review and criticism–which is definitely a form of testing.

    Again, thanks for writing.

    —Michael B.

  7. Anonymous says:

    Emotions are direct factor of state of Mind, the kind of emotions we have or feel are direct result of the state of mind we are in. The gamut of emotions is infinite so is the gamut of states of mind. Now, when sit down and write down all the emotions a person can probably have and categorize them Iam sure we’ll end up in very few (some thing like positive or negative)

    When we are conscious of our emotions and try to drill down into the root cause of emotions, we’ll find that these emotions are result of a state of mind that we have at the point in time.

    To give an example:
    A defect might look a mighty failure to test engineer, when he has no knowledge of what the product is and what the defect could mean to that product the bug might eventually end-up being a trivial bug. In this context the state of mind of the test engineer is ignorant.
    When the state of mind of the same test engineer moves from an ignorant state to knowledgeable state, the perception of the engineer would change dramatically, again the state of mind plays a role here.

    Probably if a professional can track his state of mind and the actions he does in a given state of mind, he’ll be able to understand what emotions are and how obvious there are in any field.

    And Emotions can be as tangible as any thing else, but may not be directly tangible. Emotions always results in an action which is most of the times tangible.
    The resultant actions are direct quantification of our emotions
    The result of the resultant action can be taken as an input to improve emotions

  8. Paul Gerrard says:

    Loved it. Wish I’d seen it earlier. Did you really talk over those slides at the pace they present themselves? Cool.

    Thought I’d mention a few slides I put together for a Test Management Forum session last year (completely ignorant of your pitch here).

    Slides are here…

    The slides were an intro to a facilitated discussion (hence the questions at the end). When someone gets ’emotional’ we tend to think they’re upset, angry, frustrated in some way – hence the pitch is mainly on worried test managers. (I focused on expectations. No doubt there are many other viewpoints).

    One of the big drawbacks of our industry is its domination by technology, so the industry self-selects technically-oriented people. Hence the over-reliance on technical solutions to people problems. (Prescriptive process being a classic).

    Did you take your talk any further?

    Best wishes, Paul.

Leave a Reply

Your email address will not be published. Required fields are marked *