I think this question assumes a bipolar conceptualization of affect, a perspective that is held by Russell and collegues (Russell & Carroll, 1999). They claim that our emotional experience is like a pendulum switching between positive and negative emotions (the one cannot exist in the presence of the other).
Watson et al. on the other hand, hold a different perspective, saying that positive and negative emotions can co-occur (see for example the PANAS questionnaire), since these are two independent dimensions.
I am sure you are familiar with the latest theories about the human brain's functioning.
Accordingly, the limbic system and the amigdala govern our emotions.
Our conscience has very little control when it comes to strong emotions.
Therefore I assume that a negative emotion, originated in that part of our brain will not be appeased by our conscience trying to control it by generating positive thinkings.
If you haven't yet, I strongly recommend that you read this book:
Dozier Jr, Rush W. Why we Hate. McGraw-Hill; (2003), ISBN 978-0809224791
It is probably sold out but you should find it in used books stores.
It is a must for anybody willing to enter the magnificent world of our brain and feelings.
It depends, I suppose, on your further specification of which emotions. Wolpe originally proposed reciprocal inhibition as the theory behind systematic desensitization's replacement of anxiety with relaxation. Others claimed and tested whether it was counter conditioning, habituation, or simple extinction. I'd have to go with the latter, but most attention in the literature focused on whether it worked and/or whether something worked better.
Another way of looking at this question teases apart whether a tautology is embedded. For a simple-minded example, let's consider a happy-sad scale. High and low scores on the same dimension are all that's involved. But language is slippery and can sometimes obscure the most obvious explanation. You can often find the same items in tests of different sounding constructs.
There is a theory that states humans can be in a state of mixed or dialectical emotions - Wilken and Miyamoto have a comprehensive article that covers its origins, how positive and negative emotions have been mainly measured differently, and how research has developed towards measuring emotions of opposite valence (e.g., happy and sad) at the same time can occur. They state that positive and negative emotions can co-exist in harmony and occur simultaneously, better known as the concept of dialecticism, which was first introduced by Peng and Nisbett (1999).
There's also research that suggests a correlation between positive and negative emotions. Check out Tugade and Fredrickson (2004) and their reference page for help: http://psycnet.apa.org/journals/psp/86/2/320/
Check Barbara Fredrickson's publications on positive psychology. In general, she believes that the effects of positive emotions can compensate the effects of negative emotions.
Here's the website for the organization that started the field of positive psychology. Many resources available there. http://www.positivepsychology.org/
I would recommend a look at the literature on affect, specifically on the asymmetric processing of positive and negative affect (positive-negative asymmetry). Generalizing to emotion in general, it may suggest that when positive and negative emotions are simultaneously activated, the negative would tend to overrule the positive. However the impact of the negative may be less persistent than that of the positive and the subject may engage in cold cognitive activity pertaining to the neutralization of the impact of the negative affect. I attach a chapter on the issue, and recommend more recent contributions, for instance, by Baumeister, Cacioppo, Lewicka, Taylor, and their collaborators who have approached the phenomenon from different perspectives.
I understand and respect the previous responses to the question.
You are asking for a theory. Great! That is how research begins.
First thing that comes to my mind is a question . Is this something to be researched by quantitative measures or qualitative measures?
How about a theory that considers theories of physical science? I say this because patterns and continuums are assumptions concerning the universe. My thinking might be aligned with the responses of Egon, Lumi and Guido. There are theories of positive/negative in the science of physics. You are just using psychological or affective variables. Using this theory, compare how a measure of negative emotion (Dependent variable) changes with a measure of a positive emotion. (Independent variable). You would have to define a negative emotion (use tool that measures that) and what defines a positive emotion (use a tool to measure that, too) as suggested by the previous respondents. Depending on the method, Various statistics would reveal correlations, etc or you could use the technique of modifiying/mediating variables.
Let me now if my research sounds like it is methodologically off base. However I believe that my proposal is grist for the mill. Thank you for reading.