Quantitative and Qualitative Models of Emotions

An essential principle in REBT is that we can distinguish between two sets of clearly discernible emotional and behavioral reactions to adversity: one that we call appropriate, functional or healthy, and the other which we describe as inappropriate, disturbed, dysfunctional or unhealthy. REBT practitioners and theoreticians claim that these two different sets exist, that they can be distinguished from one another (David et al. 2004), and that they result from different types of beliefs or thinking: rational and irrational. Furthermore, we claim that people can influence their reactions (i.e. their emotions and behaviors—especially those that are unhealthy and dysfunctional) by changing their thoughts.

Over the years Ellis has articulated the difference between the two sets more clearly, seemingly coming to different conclusions. In 1962 he says about healthy and unhealthy anxiety: “…frequent, prolonged, or intense negative or self-defeating emotional states such as dysfunctional anxiety (as opposed to justified and self-preserving fear) …” (Ellis 1962, p. 333). This leaves open the possibility of viewing the difference as a quantitative one. Thirty-one years later Ellis contends: “RET takes the unusual position that dysfunctional negative feelings are qualitatively and not quantitatively different from functional or appropriate negative feelings” (Ellis and DiGiuseppe 1993, p. 472).

David et al. (2005) use the following definition in their investigation of functional and dysfunctional feelings (taken from Ellis and DiGiuseppe 1993). “Functional negative feelings refer to a negative subjective experience, rational beliefs and adaptive behavioral consequences, while dysfunctional negative feelings mean a negative subjective experience, irrational beliefs, and maladaptive behavioral consequences. Defined this way, dysfunctional negative feelings seem to correspond to clinically significant problems (e.g. anxiety, depression, anger, guilt) while functional negative feelings (e.g. concern, sadness) correspond to normal negative reactions.” Although it would be desirable to have a definition that is independent of cognition and behavior, I will use this one for the purpose of this article. In general, it is not easy to define emotions without the use of behavior and cognition (see e.g. Cornelius 1996).

Outside REBT

The position that REBT takes is rather unique. No other major theory of emotion makes a clear distinction between functional and dysfunctional feelings (David et al. 2004), (except for Acceptance and Commitment Therapy, which speaks of dirty and clean pain or suffering (Luoma et al. 2007)), nor are there any data to our knowledge, according to David et al. (2005), which confirm or refute Albert Ellis’ binary model of distress. The binary model of distress has been investigated by David et al. (2005) using the Profile of Mood States-Short Version (POMS-SV; DiLorenzo et al. 1999) to measure functional (sad) and dysfunctional (depressed) emotions. The results support Albert Ellis’ binary model of distress. In another investigation, David et al. (2002) found that people can make a distinction between functional and dysfunctional negative feelings and that the terms describing these feelings are more than mere verbal synonyms for the same emotional experience.

Outside REBT the consensus is that by and large emotions are functional and adaptive and that, according to Parrot (2001), “contradictory evidence is no longer seriously considered and the complex determinants of functionality are not fully appreciated.”

Small steps towards the creation of a binary model can be seen, but it has yet to materialize: for example, Frijda (1986) speaks about constructive reappraisal in a paragraph about regulation of emotions—a view on that comes close to changing irrational beliefs into rational ones. The next step could be to conclude that constructive reappraisal has been executed either successfully or unsuccessfully, leading to functional or dysfunctional emotions, respectively. But he does not take that step.

Brewin et al. (1996) show clearly that information about the trauma in PTSD can be viewed as represented in a dual representation system. The two different memory systems will elicit qualitatively different emotional processes.

Inside REBT

The different ways in which Ellis has described the two contrasting sets of emotions and the almost total absence of a theory about the subject outside the REBT community have contributed to a divergence of views within the REBT community. One possibility is to conceive the differences between the two sets as something quantitative. The proponents of the quantitative model argue that the inappropriate set contains stronger, more severe or ‘heavier’ types of emotions than the other set. The more functional or healthier emotional reaction is considered to be ‘lighter’, less intense, but still of the same kind. One can be very angry with somebody (dysfunctional) or annoyed or irritated (functional). These two descriptions merely differ in magnitude, one being a lighter version of the other.

There is a major problem with this view. At what point do functional, healthy emotions become dysfunctional, unhealthy ones on a quantitative scale and how does this happen? Another problem relates to the difficulty of conveying the idea that rational thoughts gradually turn into irrational ones, which we would expect them to do if they are the cause of the corresponding emotions.

The public, and therefore most clients of REBT practitioners, tend to think about emotions in the one-dimensional way described above. This means that they see no qualitative differences between the appropriate and disturbed emotions. One can be emotionally affected by the loss of a loved one. This reaction can be too strong, too much, or even too little or it can be just OK. According to this view, we should try to have emotions of a certain strength (read quantity), but not of a certain quality. There are several other problems with this view of our emotional life as we will see below.

Different Ways of Conceiving the Two Sets

In the following section, some of the ways in which REBT practitioners and theoreticians explain the relationship between the two sets of emotions are described.

1. Richard Wessler (Wessler and Wessler 1980), who had a significant impact on the first Practitioner’s Guide to RET (Walen et al. 1980), takes a typical one-dimensional view. One can be anxious on a scale from one to ten. A score of nine or even six might get one into trouble, because it is more than five, which is the critical threshold. His approach involves helping people to reduce their disturbed or inappropriate emotions from say a 9 (on a scale from 1 to 10) to an appropriate level of 2, or even 1.

2. Ellis often says to clients that things cannot be more than 100% bad, referring to their way of looking at negative life events, while these events could usually potentially be worse than they are (Ellis and Grieger 1977). When he puts it this way, he is implicitly referring to a one-dimensional continuum, which is not the same as the one-dimensionality of the appropriate and disturbed emotions mentioned earlier, but he does make some remarks in that direction.

3. Ellis distinguishes REBT from all other forms of cognitive behavioral therapies, saying that they all “put severe sadness and irritation in the same category as depression and anger” (meaning that they only differentiate quantitatively) (Ellis 1988; Backx 2000). Here he makes clear that the difference is not quantitative.

4. In the first edition of the Practitioner’s Guide to RET (Walen et al. 1980), the authors used the SUDs scale (Subjective Units of Discomfort, Wolpe 1973), which also reflects the one-dimensional view. The client is asked how much he or she feels anxious, depressed or angry on a scale from zero to a hundred, zero being totally absent and a hundred the strongest they have ever experienced that particular emotion. However, e.g. sadness and depression are not clearly distinguished on that scale.

There is in itself nothing wrong with using SUDs to get some kind of description of the emotional state of the client. Nevertheless, it is not REBT, because it leaves out a very important aspect of REBT theory, namely the fact that there are two completely different ways of reacting, caused by totally different ways of thinking (just as when one accepts a client saying: “My husband—or my wife—made me angry”—one gets an idea of the situation, but it is not consistent with the REBT approach to leave it at that).

5. The second edition of the Practitioner’s Guide to RET (Walen et al. 1992) no longer uses SUDs and holds on firmly to the separation of the two sets of emotions: appropriate and disturbed. In this edition, to help to clarify the issue, the authors added an explanation under the heading ‘A Key Concept: Disturbed versus Undisturbed Emotions’ or, as they call them, harmful versus helpful emotions (p. 90).

6. Some people, such as DiMattia (in a lecture at the Institute for RET, Haarlem, the Netherlands in September 2002) conceptualize the two sets with two parallel lines; the area between the lines is functional and the area outside the lines is dysfunctional. Emotions fluctuate in strength over time and as long as they stay within the two lines they are healthy and when they fluctuate outside the lines they become disturbed or unhealthy. In fact this is again a one-dimensional concept. See Fig. 1.

Fig. 1
figure 1

Emotions fluctuating and either staying within or crossing the border of the functional area into the dysfunctional area

Full size image

Underpinnings from REBT Theory

According to REBT theory a person can experience the two different types of emotions (functional and dysfunctional) simultaneously (Ellis and DiGiuseppe 1993). An individual may simultaneously be a bit depressed and very sad. Or a person can be very depressed and only a little bit sad. This means that any combination of the two types of emotions, at any strength, is possible. This is very different from the concept that an emotion becomes dysfunctional at a certain level and remains functional under that critical point.

If we take the Wesslerian view of the relationship between the two sets of emotions, we can see that these two views are incompatible. Going along with that view, we would propose the following. Let us say that the one-dimensional emotional range goes from 0 to 10. In the middle, say at five, is the critical point. Then we consider an emotion with a strength from 0 to 4.9 functional and healthy (e.g. sadness after experiencing a loss), whereas from 5.0 to 10 would be considered dysfunctional (e.g. depression after the same event). We can plot the four basic emotions on the following diagram Fig. 2.

Fig. 2
figure 2

Functional and corresponding dysfunctional negative emotions on a 0–10 scale

Full size image

Such an approach becomes problematic when we want to include possible combinations of functional and dysfunctional emotions simultaneously. Somebody cannot, for example, be depressed (2 on a scale from 0 to 10), while at the same time being sad (3.5 on a scale from 0 to 10). A very important therapeutic aspect of the REBT approach is that people are helped to reduce the unhealthy part of their reaction to adversity but, at the same time, retain healthy and unavoidable aspects of their emotions (see e.g. the clinical implications of David et al.’s (2005) research). The conclusion must be that models such as these are incompatible with the main tenets of REBT.

Conclusion: Functional and Corresponding Dysfunctional Emotions are Independent

In summary, we can say that the two types of (negative) emotions or reactions to negative life events must be seen as independent of each other—or at least co-existing—and that the quantity of the one is independent of the quantity of the other. This is not to say that in practice we will often see them converge: the stronger the healthy emotion, the greater the chance of a strong accompanying unhealthy emotion. Another way to put this is that the two sets differ qualitatively. We are speaking about two different entities here and not about one being a stronger version, or an escalated version of the other, in the same way as irrational beliefs are qualitatively different from rational ones (the first containing demands and commands, while the second contains only desires) (see e.g. Ellis and DiGiuseppe 1993).

This can be depicted mathematically on a graph by using perpendicular axes for the two types of negative emotions: one axis (the horizontal X-axis) depicts the strength of the functional or appropriate (negative) emotion, while the second (the vertical Y-axis) is used to depict the strength of the dysfunctional, inappropriate or disturbed (negative) emotion, which operates independently of the first one (see Fig. 3). The two can exist side by side in the same way as the corresponding rational and irrational thoughts can coexist.

Fig. 3
figure 3

A negative emotion as a score in terms of functional and dysfunctional values

Full size image

What can be Explained with this Graph?

  1. 1.

    By using of this graph we can depict any combination in the strength of functional and dysfunctional emotions. So, in the example in Fig. 3, the score (X) represents a mild strength of dysfunctional emotion of some sort (whether anger, depression, anxiety or guilt) combined with a moderate strength of the functional counterpart of that emotion (annoyance, sadness, concern, or regret). In Fig. 3 we can see that the client has mainly reacted in a healthy way to a negative event. We can use the whole X–Y space to represent different scores, showing negative emotional reactions to adversities as well as the ratio of dysfunctionality to functionality. We are no longer restricted from the use of different strengths of emotions originating from the two different sets.

  2. 2.

    The graph shows that every emotional reaction to a negative event has in fact two dimensions: a level or quantity of functional emotion and a degree of dysfunctional emotion (Fig. 3). When the functional part is stronger—as in Fig. 3—we expect the person to act, think and feel mainly functionally, although there is also a (weaker) tendency to act, think and feel dysfunctionally.

Another way to look at these two aspects of a single score in the graph is to say that it is at a certain distance from the origin (0) and at a certain angle, which can be defined in degrees (see Fig. 4). Instead of looking at the actual scores on the X and Y axis, we describe the score X by the length of the arrow (vector) OX, and its direction, which can be expressed as the number of degrees in the angle that the line OX makes with the positive X axis.

Fig. 4
figure 4

A negative emotion as a score of a qualitative (angle α) and a quantitative (distance OX) value

Full size image

Now we’re no longer talking about the degree of functional or dysfunctional emotion in the score X, but about:

  1. (1)

    The magnitude of the total emotion (the combination of functional and dysfunctional components) expressed in the length OX and

  2. (2)

    The direction of that emotion compared to the X axis expressed in degrees.

The first is clearly a quantitative dimension, while the second is a qualitative one, because it tells us—in addition to how strong the emotion is—how much more or less it is dysfunctional than functional. The more horizontal the direction of the emotion is, the more functional it is and the more its energy or force is related to reality (i.e. somebody is using the energy of the emotion to influence reality). The more vertical the direction of the arrow (or vector) is, the more its energy or force is directed against oneself. That means that one does things against one’s own interests—directly or indirectly—by hurting oneself or others. Or one suffers the experience of unhealthy emotions such as rage, depression, guilt or anxiety.

3. The third point the graph explains is why people (wrongly) think that REBT promotes emotional flatness and indifference or a stoic attitude. If people think of negative emotions as one-dimensional, i.e. only varying quantitatively (the magnitude or length in Fig. 4), the only way to change the emotion is to change its magnitude. The ideal situation, according to that model, is to bring the magnitude back to zero (see Fig. 5). This figure shows how several of someone’s negative emotional reactions over a period of time could be scored. Most of the scores are more or less an equal mixture of functional and dysfunctional emotions, and there are some extremes (outliers). It could very well be that for another individual, the scores over a period of time would consist of more dysfunctional or more functional values, whereas they are arbitrarily shown here roughly in the middle (at a 45-degree angle).

Fig. 5
figure 5

A set of negative emotions and the direction to change them according to non-REBT approaches, i.e. the quantitative change

Full size image

If we only change the quantity or magnitude of the negative emotion, we flatten it out (in Fig. 5, this is done with the largest score) and are left with indifference (= magnitude 0). This leaves out the most important tool in the REBT intervention tool kit: the ability to change the quality of a negative emotion. This is the basis of REBT philosophy, that regardless of how negative events are, we always have the option to react not in a destructive way, but constructively, with healthy negative emotions.

4. What REBT tries to accomplish in relation to negative emotional reactions is clearly shown by this graph: it is not the magnitude or quantity of the emotion that needs to be changed, but rather its quality. In the graph this means not to change the length of the vector (the arrow OX), but rather its direction. The more horizontal the line OX is, the healthier the reaction will be. By doing that, the emotional reaction of the person is not flattened, but the quality of the emotion is influenced: it becomes less destructive or self-destructive, more constructive and more likely to influence reality. In general, we can say it is more (constructively) engaged with reality. This is the opposite of emotions that are more vertically oriented, which are (destructively) directed against oneself and/or others. In Fig. 6, we can see what happens with the same score X from Fig. 5. Now the magnitude of the negative emotion is unchanged, but the direction is moved towards the X axis. We would like to do the same with the other scores in Fig. 6, i.e. shift them all towards the X axis. In practice, of course, we will usually only succeed in helping people to do this to a certain extent.

Fig. 6
figure 6

A set of negative emotions and the direction to change them according to REBT, i.e. the qualitative change

Full size image

5. The graph also explains why some clients experience their negative emotions disappearing or flattening after they have disputed their irrational beliefs. In such a case they have not only reduced their dysfunctional negative emotions, but also their functional negative ones. They might say, for instance, “I really don’t give a shit anymore about losing that job!” instead of: “I absolutely don’t HAVE to get that job, although I wanted it very much and, now I know I didn’t get it, I can live with that.” Partly this is the result of an inappropriate way of disputing (replacing the irrational belief with the wrong belief, that leaves out the original desire) and partly because people simply have no other script. To put this differently: they simply cannot imagine that it’s possible to react appropriately to a negative event. This also very often means that people have the wrong notion about negative emotions to begin with, namely that both unhealthy and healthy ones are inappropriate.

The conclusion may be that people have to be trained to change their negative emotions only qualitatively, and not quantitatively. Cognitively this means that they evaluate the adversity just as negatively as before but do not rely on a demanding attitude, but rather on a desiring/striving attitude.

6. This also shows that when we guide an (emotional) reaction toward a more horizontal position in the graph (see Fig. 6) and we do not change its magnitude, this implies that its score on the horizontal (functional) axis may increase, while on the vertical axis it may decrease. When somebody experiences the loss of a partner and is both very depressed and very sad, this implies that, as depression decreases, sadness will increase. So here depression works partly as a cushion against reality, protecting the individual against a grim reality.

7. Clients can be shown that only a certain proportion of their reaction to a negative event causes them suffering, namely only the ‘vertical’ part. They can keep the healthy part which they have often tried so hard to eradicate before. By doing this, they wasted a lot of energy that would otherwise have been available for attacking their irrational beliefs. So they can now focus on the dysfunctional part of their reactions, which is much more efficient. Clearly some clients profit from this possibility, as a client explained in an interview on video after a demonstration session with Albert Ellis (Backx 2000).

8. Sometimes clients say that they do not want to give up their demand because they need it for motivation. This is another way in which clients confuse the healthy, rational and the unhealthy, irrational parts of their feeling and thinking as if they are one and the same. With the help of the graph we can show them (and literally point out to them) that the painful and counteracting part of their feelings, thoughts and behaviors falls along the Y-axis and the healthy, constructive part falls along the X-axis. Instead of believing that the one can not go without the other, they can learn to distinguish keenly between them.

Implications for the Effective New Thoughts (E): How to Formulate E According to the Graph

The reason that irrational beliefs exist is that there are rational beliefs or, to put it another way, that people have desires. We start desiring things, but then transform our desires into demands on the world. This is something we are not aware of and in particular we are not aware of the existence of the demands or irrational beliefs in our desires. But we do know our wishes and desires very well and we consider the combination of our rational and irrational thinking to be our desire. So when clients are asked to give up only the irrational component, they tend to believe that we are asking them to give up all of their desire and consequently they do not want to. Therefore it is good practice when we formulate an effective new thought (E) or rational belief for a client, to strongly emphasize the original desire in that new thought.

Let us take, for example, the common irrational belief “The world simply MUST be different from what it is right now,” referring to something someone insists should be different from what it is now. We start the effective new thought (rational counter-belief) by opposing the demand as strongly as we can by saying: “For various reasons the world has to be exactly the way it is.”

If we leave the effective new thought (E) like this, we run the risk that the client will eradicate all the emotions that he or she experienced in the negative situation: both the healthy and the unhealthy ones. When we look at the graph, this means trying to bring the emotion back to the origin (0) (see Fig. 5). To avoid this happening, we can add: “… although I would very much like the world to be different from what it is right now. So I am going to try hard to change it.”

Now we have boosted the original desire and we have avoided it disappearing together with the irrational demand. To say the same things in terms of the graph, we have first brought down the vertical part of the score as far as we could and we have subsequently strengthened the horizontal part. We try to keep the emotion’s strength (the length of the arrow OX in Fig. 4) in order to preserve its quantity, but we change its direction, or quality. We have thus moved the force of the emotion in the direction of reality and avoided the whole emotion disappearing, which would leave the client without any desires.

But the original desire, which we have just strengthened, was the very reason for the existence of the irrational belief. The desire was for various reasons transformed into an irrational demand (see e.g. Backx 2003). So by boosting the desire we now run the risk that the irrational belief will re-emerge. Referring to the graph: we run the risk that the vertical score will go up again and the force of the emotion will also be directed against the client him or herself, because we humans have a natural tendency to move a horizontal score in a vertical direction.

To counteract such an effect, we add to the effective new thought: “…, but if I nevertheless fail to succeed in changing that aspect of the world, then I will accept that and I can live with it.” Now we have attacked the irrational belief for the second time right at the very moment when it might have emerged again, because of the existence of a strong desire. In terms of the graph, we have again prevented the score from growing vertically i.e. into a demand that acts against the individual concerned.

When I offer this new effective thought to the client, I explain that these two sentences are a mouthful and it’s not necessary to verbalize all of it all of the time. Just a small part of it will suffice. But in the same way that a broad-spectrum antibiotic attacks many kinds of bacteria, this covers all possible irrational thoughts that may come up. So it’s wise to keep it at hand.

Example from Therapy

A 43-year-old male had great difficulty controlling his anger about an injustice he had experienced. His boss had betrayed him and, as a consequence, he lost his job. After strongly disputing his irrational belief that his boss SHOULD not have behaved this way, he experienced a dramatic reduction in his anger. He was very glad about that and reported that he was feeling much better as a result. But he said he now had another problem: not only did he not experience anger about the betrayal; he experienced no other emotion either. He asked the therapist: “Is the purpose of RET to let these people get away with such wrongdoing? Do I just have to let this happen?”

Clearly he felt a kind of indifference instead of a (preferred) strong negative healthy and constructive emotion about the betrayal. The therapist explained to him the difference between emotional indifference and a useful constructive (appropriate) but negative emotional reaction to negative events.

He also explained that he had a choice: he could either think that his boss had the option to betray him or even that he had the right to do so, but that he could still judge that event very negatively and try to change what could be changed. But his irrational thinking was directed towards the impossibility of the boss changing his behavior as such, which made the client very angry. A better emotional reaction would have been to feel very disappointed and irritated about the betrayal (i.e. display an accepting attitude), while continuing to view the betrayal as very negative and feeling and acting accordingly.

The therapist used the graph shown in Figs. 5 and 6 to explain that the client was not only trying to get rid of a dysfunctional emotion, but also suppressing a functional, healthy response at the same time.

  • T. So, Jim, how did you reduce your anger about the betrayal of your boss? And how did you become indifferent to that betrayal?

  • C. I told myself that what he did doesn’t matter and that he has the right to betray me.

  • T. But it does matter, Jim. You’re right that your boss has the option to betray you—as all human beings have—but he put your job on the line in the process. And that’s bad—perhaps very bad.

  • C. But when I think that, I’ll get angry again and I don’t want to be angry. So I tell myself that it doesn’t matter.

  • T. By saying ‘it doesn’t matter’ you become indifferent. That’s true. But you don’t get angry when you think it’s bad or very bad that he did that to you. You get angry when you think that he absolutely should not have done such a bad thing.

  • C. Yes.

  • T. Let’s take a look at this graph.

  • The therapist explains the graph as described above.

  • C. So you say that I can stop thinking about what my boss ‘must’ or ‘should’ do, but keep my preferences at the same time. How do I discriminate between these two thoughts?

  • T. Well, you said that you get angry when you demand a change in his behavior. But how do you feel when you recognize that he may have had all kinds of (good or bad) reasons for doing what he did, but that you nevertheless very much dislike it and also wish fervently that he hadn’t done it?

  • C. Something like “so what”?

  • T. Say the new thought to yourself!

  • The client repeats the E (effective new thought) out loud to himself.

  • C. Well, what I feel now is more like being irritated.

  • T. So can you distinguish between these two emotions: anger and irritation?

  • C. Yes it feels quite different. It still feels bad, but I can let it go.

  • T. So, instead of “so what”, what leads to indifference, you could say now “tough shit!” about the situation.

  • C. Yes, I see.

Conclusions

In this article the views of REBT practitioners and theoreticians on the relationship between healthy or functional emotions on the one hand and unhealthy dysfunctional emotions on the other are discussed. Identifying the qualitative difference between the two sets of emotions is shown to best correspond to REBT principles. This approach shows that several phenomena can be explained by this model of emotions, for example the critique that REBT tries to rule out all emotions as well as the difficulties clients encounter when trying to change their emotions.

The article shows how this model can be depicted by a graph, from which the characteristics of the emotions, in the face of adversity, can be deduced fairly easily. It shows how the graph can be used to explain the way emotions can change (according to REBT) to clients and novice practitioners.

Effective new thoughts can be constructed with the aid of the graph by reducing the dysfunctional part but keeping or even strengthening the functional part. The article ends with a therapy vignette which shows how a client can be helped, again using this graph.

A fruitful direction for research could be to examine whether the ratio of unhealthy and healthy part of emotions actually can be changed, as the graph suggests using the instruments David used in his research (David et al. 2005). But it would also be interesting to carry out research on whether explaining this model and using the graph actually helps clients to more effectively reduce their unhealthy reactions.