A Brand New Razor

You’ve heard of Occam’s Razor. According to William of Ockham (c. 1340), given a choice between two equally accurate models (of reality), the simpler model is preferred. 

Occam gives disheveled, often bloated philosophy a good shave. But not good enough. Occam leaves stubble. I propose to give philosophy an even closer shave using a brand new ‘ATM Razor’: 

“When a phenomenon can be modeled by two or more equally elegant and spare (Occam) hypotheses, the model that posits (or requires) the fewest fixed entities is to be preferred.” 

All models distort the events they model. That’s the nature of modeling. But using the ATM Razor we can minimize the pressure exerted on process in the modeling process.

All living Indo-European languages break down the perceptual field according to a more or less common ontological  grid. In English, we talk about nouns (including pronouns) and verbs and the adjectives and adverbs that qualify them.   

20th century British philosopher Alfred North Whitehead characterized all western philosophy by lifting a single verse from an Anglican hymn: “Abide with me, fast falls the eventide” – the interplay of stability and flux. 

In so doing, Whitehead was doing no more than sampling Parmenides, the father of western philosophy, who spoke 2,500 years ago of two realms of Being: Aletheia (Truth) and Doxa (Appearance).  

Being in the realm of Aletheia is monotone, motionless, unchanging and without space or time. Everything else is located in the realm of Doxa: qualities, motions, changes, and spatiotemporal relations. 

Whitehead has contributed many important ideas to western philosophy. His ‘Fallacy of Misplaced Concreteness’ is a philosophical reformulation of the first commandment of the Decalogue, the prohibition against Idolatry. Put simply, whenever human beings break down a perceptual field, there is a danger that they will over-seed those fields with nouns: ergo “misplaced concreteness."

Example: as we ‘watch’ the process of evaporation and precipitation, we ‘see’ clouds of water vapor, but we call them ”ice cream castles in the air.” That is a paradigmatic example of misplaced concreteness, our tendency to over-nominalize.  

Nouns stick…and ultimately stink. Verbs don’t. Once you’ve named a phenomenon, you have conferred upon it a kind of immortality. In many cultures, the name (noun) of something is an acceptable substitute for the thing (or person) itself.

Absent the application of some outside force, we expect our ‘nouns’ to endure...and rot. 

Want proof? Go to bed! After a bit of tossing and turning you’ll find yourself in a world where ‘named entities’ are much less durable.

We don’t expect the people, places, and things we encounter in our dreams to maintain a fixed identity or even to endure once our imagination (attention) has moved on. In the dreamworld, at least, unobserved trees do not fall in unobserved forests and if they do, they don’t make unheard sounds.

Dreamland is a cool place to visit for a few hours every evening…but I wouldn’t want to live there. If I did, you’d medicate me… or lock me up. Apparently, society cannot tolerate the complete dissolution of our nominals (i.e., the smashing of our pumpkins, aka idols). 

Verbs, on the other hand, are more forgiving. One size fits one! There’s a process and we ‘name it’ (or describe it) using a ‘verb’; but we don’t automatically assume (1) that that process will continue after we’ve moved on or (2) that we will encounter that exact same process elsewhere in other contexts.  

When we use verbs to model process, we are being more ‘data dependent’ (in the words of the Federal Reserve); we are being more faithful to experience. When we use nouns, we force ourselves to fit our perceptions into a fixed, enduring structure. We turn events into objects so we can better manipulate them.

Mapping the perceptual field onto a Cartesian grid, we can call one axis ‘nominal’ (say, X) and the other axis ‘verbal’ (say, Y). We are interested in evaluating various models of P (‘perceived process’) where X and Y are variables, but the sum of X and Y is always constant. More X less Y, and vice versa.

In such a world, absent other, overriding considerations, we should always favor the model that gives us the largest Y-value (and, therefore, the smallest X-Value). That’s our new ATM Razor. Any questions?