Imitation Done Right

A post I wrote a few years ago about the difference between invention, innovation, and imitation has received a significant amount of attention, particularly in the negative light in which I cast the practice of imitation.   And rightly so, as I was rather bold in characterizing imitation as

"an individual of inferior intelligence or insight who is merely copying a pattern of behavior he doesn't understand and putting blind faith in his ability to achieve similar results."

Having been chastised for making that statement, and having meditated a bit more on the matter, it seems necessary to qualify that - but only by inserting the word "often." Most acts of imitation I have witnessed or heard about are indeed brainless mimicry.  But this does not rule out the possibility that imitation can be done in a well-considered manner and can have positive results - but to imitate effectively requires a considerable amount of diligence.


Step One: Imitation as the Basis of Learning 

Imitation is often the first step of human learning: an infant mimics the sound of human voices in order to learn to speak.  Parents teach their children to speak words simply by stating them and expecting the infant to copy - which the infant readily does when it reaches the stage of development at which motor coordination and cognitive development permit.

It is by this method that all human beings learn to speak - by first mindlessly copying the sounds we hear and uttering combinations of we don't understand, but to which others seem to respond.  It is not until later that we associate words to the things they represent and begin to form our own sentences.   An adult can learn foreign languages by cultural immersion (rather than classroom training) in much the same way - by hearing words we do not understand, repeating phrases to learn to make the proper sounds, and eventually understanding how to compose original phrases.

But before mimicry gives way to understanding, it remains a mindless practice of merely copying something we do not understand - which serves little purpose.   And this is the reason I had been so negative on the notion of imitation in the previous meditation: many people claim to be "strategic thinkers" who are merely imitating what they see without understanding it to any appreciable degree - they have taken the first step on the journey to intelligence, but seldom take the second, or any other.


Step Two: How Imitation Yields Knowledge 

Imitation to be strategic: we should not only observe the behavior, but also the results it is achieving, and then apply our critical and analytic faculties to determine whether there exists a causal relationship between the behavior and those results - and we must be diligent in our scrutiny.   This step is often missed, or consciously skipped because it requires a great deal of mental effort to accomplish, as well as the ability to reason things that are not immediately perceptible.

Many seemingly intelligent people assume that correlation implies causation without due analysis.   Superstition and luck are based on such poor reasoning: if a salesman was wearing a particular necktie when he closed two major deals, he begins to regard it as his "lucky" necktie and puts faith in the notion that wearing that particular tie has some contribution to his success in closing the deals.   It's a rather silly notion that one has anything to do with the other, but it is rather astounding how widespread this kind of sloppy thinking happens to be.

I cannot count the number of times in which a discussion about website design degenerated into this sort of thinking: suggestions such as "let us copy the fonts and colors used on Amazon" or "we should copy the page layout of Expedia" - as if the font, color, and layout had anything to do with their success.   It's no sillier than having a lucky tie - and perhaps it is even sillier because the connection between behavior and results is not known (the salesman with the lucky tie can at least cite two instances in which the tie was correlated to success in his personal experience).

What should be obvious is that if our aim is to achieve a given result, we must not merely witness behavior that coincides with the result, but apply our critical minds to determine if the behavior caused the result to occur.  I feel like I'm being a bit punctilious here - but given that this step is so often ignored or skipped, it seems necessary to be meticulous in making that point.

The sad truth is that sometimes imitation works out for the better even when we did not do this analysis - which leads to the notion that analysis is not necessary.   It's the same lucky-tie conclusion that what seemed to be necessary (or in this case unnecessary) is in fact so, and no further thought is necessary when imitating - even when we are imitating a practice of imitation.

And so, for imitation to yield knowledge - to say in fact that adopting a given behavior will achieve a specific result - requires a more significant amount of observation, analysis, and reasoning than merely noticing that the two happen to coincide.  And this must be done with objectivity and with a careful effort to avoid many common fallacies about causation.   But the process does not stop here.


Step Three: How Imitation (Can) Yield Results

At this point in the process, it has been observed that a given behavior coincides with a given result, and a careful analysis of causality has been performed to ensure that there is more than just coincidence, such that we have a plausible case for a cause-and-effect relationship between the behavior and the results.   But before confidently choosing to imitate behavior, one step remains: a comparative analysis to ensure that the situation in which we mean to implement the behavior is sufficiently similar to the situation in which the behavior caused the outcome.

To begin with a ludicrous example: it can be observed that high-octane gasoline makes a race car go faster - and those who understand the mechanics of internal-combustion engines can explain the reasons why it is so, establishing a causal relation between high-octane gasoline and speed.  An astoundingly simple-minded person might them jump to the conclusion that he ought to force-feed high-octane gasoline to a race horse to improve its speed.   I expect that I do not need to explain why this imitation will not work.

But then consider how many sites covet Amazon's one-click ordering process: the brilliant simplicity of being able to store payment and delivery options to give shoppers the ability, with a single click, to purchase an item.   It works wonderfully for items that have no configuration options, but even Amazon doesn't use one-click ordering for items that have to be configured (a shirt for which a person must specify size and color cannot be ordered with a single click) - and in those instances a one-click ordering system does more harm than good (many product returns from disgruntled customers who received items that were of no use because they were not configured properly).

Flaws in the comparative analysis can be very subtle.   Tradition and documented procedure often suffer from the failure to consider if the present situation is analogous to a past situation in which a given action had a given result.   As such the notion of consistency is flawed by its assumption that the present is exactly like the past in all the ways that matter and that past practices can simply be imitated to achieve the same results.

And yet, discussions about selling complex products such as investments portfolios and insurance policies invariably involve at least one person who claims "we should do one-click ordering, like Amazon," after which at least one other person must carefully (and often repeatedly) explain what should be painfully obvious: that one-click ordering isn't even a viable option for selling complex products.


Summary: Imitation Done Right

To boil all of this rumination down to a simple procedure, strategic imitation consists of three steps:

1. Observation - Witnessing the coincidence of a behavior and an outcome
2. Causal Analysis - Determining if there is a cause-and-effect relationship between the behavior and the outcome
3. Comparative Analysis - Determining if the cause-and-effect relationship will remain valid an effective under other circumstances

Skip any of these steps, and imitation becomes mindless and ineffective.