Manual Testing Deniers

There are groups of testers out there right now denying that the term “manual testing” — or the reality of the term — does exist or has ever existed. To me this is a bit of historical revisionism. Let’s talk about this.

Being a little tongue-in-cheek here, when I hear these discussions I feel like I’m dealing with the evolution deniers or those who deny we ever landed on the moon. It’s not quite that bad, but I do think there’s a problem here.

Cases In Point

Consider the article Manual Testing Never Existed. Saying something never existed is an ontological statement denying that a particular part of reality was ever, in fact, real. Consider this from the article:

“I absolutely disagree with the existence and the usefulness of the term ‘manual testing’.”

It’s one thing to take issue with the usefulness of the term. I often take issue with it as well. But it’s another thing entirely to deny its existence.

Conflating those two aspects — existence and usefulness — is interesting because if the concept didn’t exist, how could you determine whether it was useful or not? Ironically perhaps, part of what good testers — as a manual (human) activity — do is look for inconsistency or contradiction. Yet our writings often get permeated with those very things when we get in denial mode.

Consider this test talk. A quote from it:

“There’s no manual testing! It’s like being afraid of werewolves. Manual testing doesn’t exist in exactly the same way that manual research doesn’t exist.”

The problem here, in both cases, is that arguing by negation is something that has proven to be untenable in the wider industry. The more testers argue “there is no such thing as manual testing,” the more people tune out. Except the testers who are already predisposed to believe this, which isn’t the audience we need to reach.

Manual Testing Did (and Does) Exist

There has been a concept of manual testing since that term was associated with human debugging decades ago and led to the rise of better debuggers. Whether it should have been called that is a different point from the fact that it was and is called that.

Acting like the term has not been a reality for a long time or acting like it still isn’t a reality is almost ludicrous. About as ludicrous as comparing it to a fear of werewolves. Testers already have a shaky reputation in some spots; acting as if they are denying reality doesn’t help that. Equating someone’s notion of an idea (manual testing) with a — perhaps? — mythical concept (werewolves) is dismissive, at best.

Yet … I Agree; Up to a Point

I want to make one point very clear here. I do happen to agree that the term “manual testing” is often non-indicative of the true scope of what testing is and should be. I do agree that, in some cases, the term can be equated with “mere” manual labor, to be done away with.

So what I try to do is this: rather than argue from a negative, I argue from the positive and say what testing is. I talk about how testing is a design activity, an execution activity, and a framing activity. Each of those activities operate at different levels of abstraction and can utilize different heuristics and different techniques. And, yes, some of those are done in what we traditionally call “manual” fashion: by a human being.

This way of discussing it has the benefit of matching up with just about every other discipline, particularly scientific ones, where the term “testing” is used. It also starts to reframe the discussion, which I think is necessary more than ever.

Even beyond a tester’s reputation, denying an existing nomenclature that is entrenched rarely works. I do agree with challenging assumptions and terminology, of course. The trick is that a lot of times that challenge has to be presented as a reframing — and one that actually matters to what people value. Otherwise discussions can come off as some internal debate within a discipline that has no merits to anyone else.

Does the Wording Matter?

Does it actually matter if part of the practice is called “manual testing”?

As I said above, many people, sometimes correctly, see this as being a way to get rid of “manual labor” and replace it with automation. Here’s the problem: those same people probably want to do that even if you switched the term to “human testing.” They’re not confused about whether “manual” means “human.” They’re concerned that the human is inefficient and automation is better.

Adding in the distinction between “testing” and “checking” rarely matters. In those cases, those people who over-rely on automation still want to automate as much as possible and the nuance of the distinction is often lost because it simply doesn’t matter to them.

Okay, yes, fine: but should it matter? Let’s consider a few examples.

  • Astronomy. As many know, Pluto was demoted from being a planet to being a dwarf object. Yet whether Pluto is or isn’t a “planet” — by that name — matters a lot to some people. Absolutely not at all to others. The vast majority of what we know about Pluto and how we study it changes not at all with this demotion.
  • Paleontology. Whether the fact that what we called a Brontosaurus is really, in fact, an Apatosaurus can be largely meaningless to anyone who doesn’t care about the details of dinosaurs. What we study about this particular creature is a separate matter from what it happen to be called.

I’m honestly not trivializing here; I’m just showing that what can seem of supreme importance within a discipline can seem like so much semantic nitpicking to everyone outside of that discipline, particularly when what they need is not relevant to the distinction.

And, keep in mind, it’s usually those outside the discipline we most need to convince. If you’re trying to get funding for your astronomical observatory, going on about how “Pluto is not actually a planet and never should have been called one” is not going to make your case. Getting funding for your next digging expedition is not going to go as well if you focus largely on how the Brontosaurus never existed. (By the way, the Brontosaurus is back.)

That’s why I would rather give people the vocabulary, the nuance behind the vocabulary, and then show how that vocabulary can be framed in discussions where it doesn’t sound defensive or in denial of a reality that not only do others perceive but that is, in fact, reality.

But shouldn’t our vocabulary be consistent with the discipline?

But No Other Discipline Uses “Manual”!

Testing, as a discipline, exists behind just about every human endeavor that requires investigation, experimentation, and exploration. Whether that be physics, biology, teaching, molecular chemistry, archaeology, and so on.

And, yes, while we don’t speak of “manual programmers” in our context, the term “manual” does come up in those scientific disciplines many times. That test talk article I quoted says this:

“Nobody talks about manual medicine.”

Well, actually, they do! “Manual medicine” is known as chirotherapy, from a Greek word meaning “hand.” Okay, I’m being purposely obtuse there just to make a point. I say that because it is true that we don’t talk about “manual doctoring,” as an example. Yet the term “manual” does come up a lot in various disciplines.

  • “We’re doing a manual landing.” (avionics)
  • “We have to do manual adjustment of the particle collimating mirrors.” (particle physics)
  • “We have to dig the overburden manually.” (archaeology)
  • “We have to grade the essay tests manually.” (teaching)
  • “We’re doing manual extraction from the lexicostatistical wordlists.” (phylogenetic reconstruction)
  • “We manually select the operational taxonomic units that have the largest number of derived evolutionary steps.” (plant taxonomy)

Now, if you look for the term “manual testing” for those areas, I agree: you won’t find it as much. But you might very well find the term “manual procedures.”

And the reason I belabor this point is because nobody — literally nobody — gets confused about a very simple fact here. What the use of “manual” means in those contexts is that human action and thought is required.

In those cases, and in those disciplines, we don’t reframe away from the human as has happened with the technology sphere and, then, only among testers. And even then usually only among the paid consultants. Most of the time you hear this denial of manual testing, you’ll be able to trace it back to someone who heard it directly from a paid consultant.

Regardless of the extent to which you agree or disagree with me here, I hope I’ve at least shown that all of the above can be a way to start drawing parallels between disciplines but also showcase some of the real differences, which lead into better discussions about the power of “manual (human) testing.”

Better Diagnostic Reasoning

Humans can reason predictively (from cause to effect) or they can try to reason diagnostically (from effect to cause). Humans are notoriously worse at the latter than they are at the former. That’s often what we see here. People see an effect (automation) and they reason back to a cause (the use of the term manual testing) and treat that, partly, as a causation.

Actually, I’m being a little simplistic there but I think it’s demonstrable that many of the reactions against the term manual testing are cast in a diagnostic format. To provide a diagnostic counter-argument to that, I wrote about the arguments around “the death of manual testing” before and pointed to where some of these perceptions may be coming from.

I think the first article I quoted earlier (which is titled, in full, “Manual testing has no future, manual testing has no past. It never existed!”) does bring up very good points about what testing is as well. Those points alone did not need to be couched, however, in the context of manual testing denial, which can hurt the message and even preclude many from reading it in the first place.

So, yes, I do believe that the reaction against the term “manual testing” is often a reaction against the term “automated testing” and the fear that it is somehow diminishing the industry or removing the need for humans. That’s not true. The technocrats are doing that. And the managers who listen to the technocrats are making that possible.

But there are much better ways to combat this than sitting around denying the current reality of the term manual testing or making ontological statements that deny that the term “manual testing” exists or ever existed.


About Jeff Nyman

Anything I put here is an approximation of the truth. You're getting a particular view of myself ... and it's the view I'm choosing to present to you. If you've never met me before in person, please realize I'm not the same in person as I am in writing. That's because I can only put part of myself down into words. If you have met me before in person then I'd ask you to consider that the view you've formed that way and the view you come to by reading what I say here may, in fact, both be true. I'd advise that you not automatically discard either viewpoint when they conflict or accept either as truth when they agree.
This entry was posted in Testing. Bookmark the permalink.

One Response to Manual Testing Deniers

  1. Jim Hazen says:

    Great post Jeff!  As I am “manually” typing in this response I see that again we in the Testing community are our own worst enemy when it comes to vocabulary and meaning/context of words.  The word “manual” has come to be symbolized as being bad and with low regard, and that is due to the emphasis upon automation.  Which in itself is a misnomer in meaning.  Automated Testing is not automating the act of testing, it is mechanizing the process of execution of a test.  There are things that have to be done before, during and after for an automated test to run.  There are “manual” interactions the tester needs to do.

    This reminds me of the debate of the use/meaning of the word “Bug”.  Some people prefer defect, while others like issue or anomaly or inadequacy.  All of which lead to a “functionally undesirable business application response’ or FUBAR.  This causes a situation where the systems not available for use, or SNAFU.

    This SNAFU was the real culprit behind the FUBAR and explained the inadequacy which caused the anomaly raised by the issue that was depicted by the defect which just bugged users.

    This just leads users to believe that no matter what, a ‘bug’ will eventually lead to a FUBAR that causes a SNAFU.

    Whatever, a rose by any other name.

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.