Although the breakdown of behaviour into signals, markers, tokens and handles (SMTH) is fairly straightforward, it seems to be sufficiently novel that a conceptual leap is required. Hence this page has been expanded for greater clarity. This very interesting theoretical exploration began as an addition to the page on Creative Transduction.
One aspect of SMTH, which perhaps should be emphasised more, is that these components arose from observation. This explains why there are theoretical problems still to be addressed. The scheme was not conjured out of the clouds purely hypothetically, but evolved during prolonged study. For example, the notion of the handle could have been inspired on noticing that in a very feminine environment, there is commonly reticence about using someone’s name. The distinct male and female attitudes to SMTH originally distinguished them.
A handle is a request which evokes a fixed and predetermined response.
A handle state is a condition in which handles are issued.
The basic handle state is knowing someone’s name. This enables the “basic handle” to be issued: calling that name. The fixed and predetermined response is that they respond (typically, asking why they have been called).
Definitions and a few examples of signals, markers, tokens and handles are given here.
A good illustration of handles and handle states is a parent calling in a child. Let’s say Mommy stands on the doorstep and calls in Billy. They know each others’ names, the basic handle state. They are genetically related, another handle state (“family obligations”). Being a parent is a handle state (expectations of care, obedience etc.). The parent expects to be obeyed. Billy, however, is given to disobedience.
Mommy calls Billy in from play. The handle is the command: its proper response is that Billy complies, and does what his mother says. When Billy shouts “No, I’m going to the park” the handle is mutated to a marker. If he runs away without saying anything it is the mutation of a handle to a signal – the running away is the signal. (Like all signals, it is ambiguous – Billy could be running to collect a toy before he comes in.)
By convention, we look at the game from the Protagonist’s point of view. Billy is the Opponent in this game, and he has altered his handle response, but from the perspective of the Protagonist, her handle has been answered with a marker or signal.
Very many handles are implicit. We don’t normally see signs outside houses saying “Would visitors please not drive onto the lawn, but park at the kerbside.” The request is implicit; the difficulty in this instance is precisely specifying the handle state. It will involve norms and social expectations of consideration for others.
At this point some threads can be drawn together and an important model clarified. The Mugging Game was first mooted in an earlier News Update and will be elaborated.
A man has had a merry evening down the pub and is drifting, more or less in a straight line, home. The handle which is relevant to this game is that he has an expectation of not being waylaid and robbed – such actions are explicitly discouraged. His implicit request, i.e. handle, is not to be mugged. The handle state underlying the handle is the legislation outlawing it. It could be said that the legislation is merely a formalisation of the reasonable expectations of a civilised society.
The layman might consider the game to be started when a mugger appears and waylays our reveller. After all, ‘protagonist’ can in common parlance be a synonym for ‘aggressor.’ In game theory however, the Protagonist is the main player, who starts the game, and this game began when our reveller set forth for home with the expectation of not being mugged along the way. It is he who issued the first handle.
Now a mugger appears on the scene, the Opponent in this game. He attempts to issue a second handle, demanding money and valuables. The important point, and the one given fresh emphasis, is that this handle is invalid (not ‘soundly based’). For reference we compare this situation to the formal environment of the military. A handle in the form of a military order must be obeyed unquestioningly: in this case the handle state is rank. Soldiers of higher rank issue valid orders to soldiers of subordinate rank, and the soldier who defied such an order would face punishment. If an ordinary soldier were to give an order to an officer it would be invalid, due to the absence of a handle state. In a mugging, the mugger threatens punishment on non-compliance of an invalid handle.
This raises a theoretical problem: What precisely is an invalid handle? A handle without a handle state is invalid, but what else is it?
Conceptual problems aside, in the Mugging Game the Protagonist’s handle is mutated to a marker, because the criminal invariably marks his victim. That is, the handle is the request (not to be mugged), and the Opponent, instead of giving the normal handle response (not mugging), places markers instead. The Protagonist is unambiguously involved with the Opponent; he suffers loss, distress and other consequences of the criminal act; he is marked. Inflicting such markers is part of the payoff for many criminals, especially the amateur ones – it is an expression of the criminal’s ego, much like a person leaving grafitti marks territory.
A law is an explicit handle, and it is proposed that any criminal act involves the mutation of a handle. It follows (if the proposition is true) that if no handle mutation takes place, or if the handle is invalid, with no handle state, no crime has occurred.
Then it seems that governments can issue invalid handles, particularly in creating legislation which benefits non-natives. A government’s handle state – the entire reason for its inception and existence – is that it acts in the collective interest of its native population.
The hierarchy of these components is that handles carry the most weight, followed by markers. Signals and tokens, being ambiguous, have less significance. Females favour signals and tokens because they are ambiguous (and thus manipulative), while males prefer markers and handles because they are not.
The importance and masculine nature of handles means that we can expect a rich variety of procedures to exist to weaken them.
The following may be an instance of hard-testing the theory, and finding that practical reality is complicated. Informally it might be supposed that an invalid handle is simply a marker, but we need to be more rigorous than that. The theory – that all actions can be expressed as signals, markers, tokens and handles – must withstand this speculative assault or it would itself be invalid.
Throughout the development of PA I was mindful of a famous book in mathematics called Counterexamples in Analysis. First published in the 1960s, it details scores of “almost theorems.” That is, mathematical theorems which seem generally applicable but are really not. To disprove a seemingly general theorem all it takes is a single counterexample, no matter how tortuous or convoluted, and if such an exception exists then the “theorem” is consigned to the dustbin. The book gives many examples.
The current problem can be compared to an earlier question: Is there anything which is a handle but not a marker? Handles usually take the form of signals, markers and tokens. For example, talking to someone is a marker. Asking for a loan to be repaid is a handle in that form. It could involve a token as well, if the request was symbolic of something else. The money may have been loaned as a test of honesty say, or the request for repayment could represent the finalising of relations.
What was needed to resolve this question was some scenario which involved a handle only. That is, a “distinct handle” that was a handle and nothing else. A single instance would do. It took several years, but an example was eventually found. The fact that a handle can exist in isolation proved beyond doubt that the handle is a distinct entity.
It is a powerful technique. If I were to claim that “All swans are white” all that is required is the production of a single non-white swan and the assertion is disproved. Similarly, the endeavour in this instance may amount to finding a single example, no matter how obscure.
Here are some cases involving invalid handles:
The Mugging Game above, except that our homeward reveller is not quite as drunk and vulnerable as he appears, and immediately attacks and fights off the mugger. This is a masculine response, because it reinforces a handle, which males favour.
We can refer again to the military, a masculine environment in which much behaviour is formalised. Were a private to give an order to an officer, he would certainly get a response, but it would not be the normal response to an order. He would automatically be marched off on a charge of insubordination.
Or perhaps not. So ingrained is the military ethos that to a soldier, the idea of giving a superior officer an order is practically inconceivable. Were a psychologist to attempt an experiment along these lines, he would have the utmost difficulty getting soldiers to participate. An alternative outcome is that the soldier who attempted to give an order to a superior officer would be marched off not for punishment, but for psychiatric evaluation.
In any event, the invalid handle is actually a different handle – it is a request for punishment or treatment. A token is involved: the “order” means something else. The subordinate soldier has issued a token in the manner of a handle.
Suppose it is the days of the Wild West, and imagine a town, perhaps close to the Mexican border. Bandits have been active in the area, raiding, robbing and killing as they please. A large gang turn up at this isolated town. The town’s single lawman is away or incapacitated. The gang converge on the bank, burst in and spread out until all the bank staff are covered. The leader of the gang puts a gun to the manager’s head: “Open the safe or die.”
Is there any doubt that the manager will give his own life priority over the bank’s money? That his ‘fixed and predetermined response’ will be to open the safe? (We might be reminded of Private Moyse, but Moyse died for more than mere money – he put to shame all those diplomats and politicians who kow-towed to the Chinese subsequently.)
The bandit’s order, backed by threat, is invalid by the standards of civilisation and the law, but it seems that it is a handle in the immediate moment. The handle state is the gun to the head. It would appear that in the context of immediate practical reality the handle is valid, though it is invalid in another, wider context.
Beyond that immediate moment, the bandit’s handle is also a request for something else. The gang’s activities are plainly intolerable, and responses and countermeasures are certain to be forthcoming from the bank, the government, or both.
Looking at the problem another way, using the principle of examples and counterexamples again, the problem may amount to finding an invalid handle which is not a signal, marker or token. This could be a tall order.
One possibility is to apply the earlier example of the distinct handle, with the handle being invalid this time. That distinct handle was the expectation (implicit handle) that one of a couple will think fondly of the other while they are apart. This was the long-sought example of a handle which need not involve any signal, marker or token.
What if someone were deluded, and believed that an emotional bond existed when it did not?
If this invalid handle is simply that, and nothing else, we have our distinct invalid handle. However the test fails; the request is not what it ostensibly appears, thus it is a token. In this instance the invalid handle is a token of his delusion.
Finding a distinct invalid handle really is a tall order. Indeed, properly so. Suppose a distinct invalid handle can be shown to exist. It must be neither signal, marker nor token and the handle, such as it is, is invalid. The claim is that all actions can be described using SMTH. There can be no exception. The existence of a distinct invalid handle would be fatal to the theory, because the action would be incapable of being expressed using SMTH.
I have been somewhat lax here, and not addressed the problem of invalid handles to its fullest extent. This is a speculative exploration of an invalid component, not a formal paper. Hopefully the exploration has been of interest, and may be instructive.
Further analysis of the complex interplay between handles and tokens might most readily be directed toward pricing. If a fixed price is displayed it suggests a handle; when bartering is expected the price quoted is a token. A future, perhaps less challenging, exploration could treat requests which are not handles.