Sitemap
Ranjit Randhawa

Ranjit Randhawa

7 min readJan 30, 2017

--

Positive and Negative.

Positive and Negative are words that are used, without defining the magnitude, of the underlying stimulus being applied, whether external or visceral. The trouble with these words is that there is no easy way to quantify the magnitude associated with whatever stimulus is being discussed, unless the stimuli has a physical manifestation that is. But assuming one was able to do so, one would have to assign a sign to some number representing the stimuli. We could guess the sign and magnitude by the resulting action taken by the test subject, but that defeats the purpose of our effort, the ability to predict that action.

It seem that assigning a sign of plus (+) to positive and minus (-) to negative would fit with out understanding of what these words mean. The magnitude of each influence would be different depending on the measurement standard being used and whether it applies generally to all or it is specific to an individual. Irrespective of such difficulties, it would make for rational decision making as one could then evaluate bi-valence stimuli and decide whether our action to take should be immediate and large or we can be casual about it as the possible reward or punishment is not that great or onerous respectively.

Evaluating bi-valence values is quite interesting and worth thinking about. It is a standard model for those studying the brain and psychology, but here we shall start and stay at the most basic level. Consider the numbers 5 and 3. Not having a sign in front of them we assume it is + by accepted definition, addition would give us 8 and subtraction +2 or -2 depending upon whether we subtract 3 from 5 or 5 from 3 respectively, about what normal day to day life would expect us to do and conclude. But things change when the two numbers are +5 and -3. If we add them, we get +2 and heave a sigh of relief for having remembered our bit of schooling. But what happened to the 6 that was taken out of the from the sum (8), taking the two numbers without signs, that is their absolute values. If not having thought of it before, we can conclude that there really is no 6 left as the -3 and +3 from +5 would add to be zero. Of course if we were to subtract the two bi-valence numbers, it gets a bit complicated as then one could get different answers depending upon how one sets up the subtraction.

If our two numbers were -3 and +3 to start with, summation does give us zero as per rules. We then could say that remainder was zero but the activation level was 6, an indication of the magnitude of the two numbers representing the underlying stimuli. Why call it activation? Biological cells activate based on accumulated ions, which of course do have a sign. So, for evaluation of bi-valence values, it is best to consider summation as the arithmetic operation of choice, which also meets our desire to find out the result of our evaluation of two opposite stimuli. The remainder tells us which influence is larger, while the activation value tells us how great the two stimuli are as discussed further.

We should be happy that things are as they should be. Evaluating the bi-valence stimulus of +5 and -3, the activation value 6 is important as it would indicate the magnitude of the two influences and +2, the remainder, would give us some help in deciding that maybe we should reach for that piece of candy. But consider that the stimuli were +10 and -6, that is each is twice our previous example. The remainder is +4 and the activation value of both stimuli will be 12. Now we may go for more than one piece of candy from the bowl, but still conditioned by the larger activation value being twice our original 6, maybe not more than two pieces.

The use of activation to describe the missing magnitude between the two bi-valence values is to try and make a connection to how neurons fire, that is, become active and start sending electrical pulses (spikes) to other connections. So far we have discussed numbers as single magnitudes but it is more realistic to consider a series of values that could be + or — over a certain time span. It should be possible for a neuron to figure out the periodicity of the stimulus, assuming that it was periodic, by balancing the positive values with the negative. To help, consider a sinusoidal waveform, which has one positive pulse and one balancing negative pulse. A neuron that was compatible with a given period would then fire if the stimulus had that same periodicity. The sum of positive values should equal the sum of negative ones leading to zero remainder and a calculated magnitude. More importantly, this leads us to understand how time could be inherent in biological systems. Quite an important concept as now we can assume that the many cycles that control us, can be generated in a distributed system rather than being centralized.

Getting back to bi-valence, it gets a bit more difficult in the case we have been considering, when the stimuli are +10 and -8, the remainder is +2 therefore making it attractive to go for that candy, but the stimuli activation value is 16 and therefore one is more aware of both stimuli and the evaluation calls for some attraction, but also more caution, hence maybe only half a piece of the candy. Of course there are many stimuli acting upon us, both external and visceral. To aggregate them so as to model our daily actions is quite out of the question in this discussion, but makes for interesting speculation.

Get Ranjit Randhawa’s stories in your inbox

Join Medium for free to get updates from this writer.

One could ask as to what happens when the stimuli are either only positive or only negative. In terms of our actions, the positive ones should give us the enthusiasm to help our selves to whatever is being offered and if negative, to withdraw as quickly as our legs would allow us. There would be no remainder in either case (remember. we are summing) and the stimulus numbers leading to activation of the action we should take. But there is another understanding to these such numbers, and it concerns correlation.

Correlation is another commonly used word, similar to positive and negative. All it means is that there is some relationship between two stimuli, (variables in mathematics if one cares to know) over some span of collected data. This collection could be based on time but not necessarily so. If after some number crunching, using well developed mathematical formulas, the result is +1 or -1, as it would be for our selected example of the stimuli being either all positive or all negative, we conclude that the variables are perfectly correlated. When one stimulus increases the other one does too if the correlation was +1 or this movement would be opposite if the correlation was -1.

In real life the correlation number is usually some number between +1 and -1 and if 0, there is no correlation. But the correlation number is for the entire data set, one cannot assume that at some point in the data set, that the correlation is either positive or negative. But here, the human brain is rather good at making, usually, a correct estimation of what the correlation is likely to be. Of course making a decision based on a small set of the data is fairly common in human experience, we are generally not aware of the larger data set which could have led to a different result,

It does make one wonder whether the brain actually works using bi-valence as its basis to make this estimation and if so how. In case the test data was all positive numbers, one approach would be is to think of the data in terms of looking at the difference between the consecutive data values which then would provide the necessary bi-valence values for evaluation purposes. This is different than the taking the differences between the averages of the two variables and the individual actual data values, as usually done by the standard formulas available to us.

There is one further thought to consider. If the two variables (stimuli) are periodic over the time span we are considering and opposite to each other, than over that span, the correlation will probably come out to be zero, that is no correlation. But for us, the idea that the two variables moved opposite to each other is an important piece of information, it would certainly come in handy if one was playing the stock market.

As was discussed earlier about the possibility of finding the periodicity of a stimulus from the time sampled data, it does describe what is known as auto-correlation. To take an example, consider again a sinusoidal waveform, it goes through half a cycle of positive values and the half a cycle of negative values. It seems obvious that from the sum of these values, the periodicity of the waveform could be determined. The activation value would give us some idea of the peak of the waveform. Of course it is not a big revelation, there are other more mathematically proven ways of doing the same. But they tend to be not as intuitive as imagining summing positive and negative values.

One additional connection to be considered is between physics and biology. Consider Newton’s law that for every action there is an equal and opposite re-action or words to that effect. To put it in terms of our daily understanding of human behavior, for every action taken by those in authority, there are unintended consequences, and opposite to the action taken. Of course the human experience has to be taken over a period of time, instead of the immediacy indicated by Newton’s law. From this thought, again we have to accept that our existence as a society is ruled by bi-valence rules, the only problem is that we have no easy way to quantify our stimuli or forecast what the re-action would be. If we could, our behavior could be modeled and our future easily predictable.

Here we have only talked about evaluating bi-valence values. But there is one other model to consider and that is hierarchical, a common structure among human society. Think about how evaluating bi-valence values allows for hierarchical social structures and the odd idea that there is no such thing as absolute laws to hold us together.

--

--