From Observation To Knowledge

Two weeks ago, 718,890 French 18-year olds spent four hours writing philosophical essays as part of the Bac nationwide exam they must pass to graduate High School. Among the topics offered to the humanities students was “Is observing enough to know?” They must elaborate on the meaning of the terms, argue for and against an answer, quote the philosophers they were taught, and draw conclusions. And it is written long-hand, without any illustrations or hyperlinks. I couldn’t compete with them, but this particular topic resonates with me because of my time observing factories.


Although observation is, strictly speaking, done with the eyes, we usually include perceptions through sound, touch, smell, and possibly even taste, but exclude information collected by indirect means. Your observations are taken directly from their object, in person, at its actual location.

Observing a factory requires you to walk through its shop floor. Reading documents, watching videos, or listening to presentations is not observation. Any mediated communication filters out potentially relevant elements — you can’t tell what a shop smells like from a video — and cannot avoid reflecting the biases of the intermediaries who generated the material — through omission, emphasis, or distortion.

There is, however, more to observation than raw sensory intake. It is even true of the ordinary act of seeing. Our brain doesn’t just receive signals but matches them against memorized patterns to recognize a banana, an AK-47 magazine, or a paper airplane.

Likewise, when you visit a welding shop, you automatically match it with welding shops you have seen before. The other ones were sparky and smelly, with protection against the blinding light around each workstation, human welders in full-face masks, and a cloud of fumes overhead.

In the one you are visiting today, however, you are breathing as if outside. You notice the contrast with the previous ones and it prompts you to look more closely at the ventilation system, which is more than elaborate than usual. And the next one you visit has neither sparks nor smoke and is free of the protective barriers found in earlier shops.

This prompts you to check out the technology used, and find out that it is friction-stir welding. Even finding out the name, however, is beyond observation. Just watching the process cannot tell you what it’s called, let alone its physics or economics.


Observation is necessary but, no matter how rich a store of relevant patterns you have in your mind, it generally does not tell you enough to acquire applicable knowledge. You need different levels of knowledge to lay out a welding line, program a specific operation with a robot, or design welding equipment. Whichever level is relevant to you, the information captured through your senses must be supplemented with input from experts and written materials, including literature and the data generated by the specific operations. If the knowledge you seek is the skill to execute or run the operation on a daily basis, you need to practice it until, according to TWI principles, you can explain the key points of every step as you perform it.

These considerations raise the question of what it is that we call knowledge, a topic I posted on before. I should be obvious but the use of terms like Knowledge Base for sets of rules or user forums had muddied the waters by conflating knowledge with beliefs, prejudice, and superstitions. Discussing the content of these systems, we end up using oxymorons like “false knowledge,” which belongs with “alternative facts” and “necessary waste.”

Mortimer Adler may be unknown to French teenagers but he hit the nail on the head when he wrote: “it is generally understood that those who have knowledge of anything are in possession of the truth about it.  […] The phrase ‘false knowledge’ is a contradiction in terms; ‘true knowledge’ is manifestly redundant.”

Not all knowledge is related to observation. Procedural knowledge, for example, is not. If paragraph 62.5.3 of the fire code says that two particular objects must be at least 6 feet apart, it is a rule you must comply with, independent of observations. Following the rule may actually protect you from fire, but that is a separate issue.

On the other hand, scientific or technical knowledge is linked to observation. Specifically, it is comprised of theories that are confronted with observations. Agreement of predictions with observations doesn’t prove the truth of a theory, but disagreement refutes it. And we choose to treat as true the theories that observations have failed to disprove.

The conventional wisdom of manufacturing and business does not measure up to knowledge. It is full of beliefs that are passed on through generations without being checked by observation. The Hawthorne effect, for example, is often presented as a universal phenomenon, based on research on just 5 women, 80 years ago, and with methods that wouldn’t pass muster today.

Just a week ago, a Gemba Coach post opened with “I’m not saying there are no good consultants. Of course there are; same bell curve as in every profession…” repeating an idea promoted as part rank-and-yank  management, where employee performance is made to follow a bell curve by forcing managers to “grade on the curve.” The grades then serve as a justification for firing the “bottom” 10%.

When you analyze cases where performance actually is an observable number, you don’t always find a bell curve. Looking at the total riding time for competitors in the 2012 and 2011 Tour de France, I found the performance distribution much flatter than a bell curve, perhaps because the riders were the 200 fastest cyclists in the world. But the employees at a leading company also presumably are the best in the world at what they do. If performance can be represented by a number, they are also at the upper end of its distribution, and extreme values are not usually on a bell curve.


Observing is not enough to know, and not all knowledge requires observation. Scientific and technical knowledge, however, does.

#Observation, #Knowledge, #Welding