Consider one of the artists whose work was displayed in the August issue of 'The Enneagram and the MBTI'. He is an INFJ who is also an Enneagram 4. His three highest scores, using the RHETI, are - 4, 1, and 5 - in descending order. According to the statistical data generated by the EM study, INFJs as a group are found most frequently to be 4s. The next enneazone in which they are most highly concentrated is 1. The third and final zone in which they have a concentration that is greater than one would expect if distribution were random (I=1.0), is zone 5. In other words, INFJs as a group seem to prefer enneazones in exactly the same order as this individual - 4, 1, and 5.

Is it a coincidence that this individual's top three scores (ie, his 'preferences' for the enneapoints) exactly reproduces the distribution pattern of INFJs as a group? How often does this occur? Can we measure how frequently this phenomenon occurs? What does it mean?

Section One: Statistical Generalizations vs. Individual Cases

skip to section two

We have sometimes heard it claimed that an individual's MBTI scores can be 'predicted' using the RHETI or some other enneagram instrument. Or, vice versa, that an individual's enneagram type can be predicted using his or her MBTI scores. Pat Dinkelaker and I want to state up front that we do not believe that one instrument can replace the other in this way. And it was never our purpose to replace either system with the other.

Furthermore, our interest is in understanding the relationship between the two systems, not in predicting, in the case of any given person, the individual's MBTI type given his other Enneagram type (or vice versa). This is not to say that a theory about the relationship between MBTI and Enneagram type should not be tested on the basis of how good a job it does in predicting how MBTI types will distribute across the enneagram. A good theory should help us to understand why distribution occurs as it does. But there is a difference, of course, between predicting statistical distribution patterns, and predicting the behavior of any given individual. We can, for instance, predict that in a large number of cases a tossed coin will turn up heads %50 of the time. But this tells us nothing about whether the next time we toss the coin it will turn up heads or tails. Similarly, we might be able to predict that the INFJ is more likely to score as a 4 than as any other Enneagram type. But this will not ensure that the next INFJ we come across will be a 4. And it surely does not mean that he or she 'should' be a Four! But neither does it mean that every individual is equally likely to score as any of the nine Enneagram types!

In using statistics to predict human behavior it behooves us to be careful to distinguish between the individual and the group. We must honor and expect individual differences, while trying to discern general patterns that exhibit themselves across large numbers of cases.

There is a pitfall that awaits us in our natural tendency to compare the test scores of a given individual to the statistical 'norm' - deviation from the statistical norm can easily be mistaken for moral deviation. This mistake can be even further aggrevated when individuals have Enneagram and MBTI combinations that seem, as Pat Wyman puts it, 'incompatible'. As I've intimated in the FD33 paper in this issue, it may not always be the case that such a seeming contradiction reflects conflict within the individual - it could also be due to the fact that the MBTI, due to limiting theoretical assumptions, renders false conclusions about the 'type' of individuals who have non-traditional preference orders. Or to the fact that the Enneagram manifests 'S-N blindness'.

A good theory should strive to explain why some individuals don't have combinations of Enneagram and MBTI type that are more typical, and understandable. The model that Pat Dinkelaker and I have proposed is fairly successful in understanding the general distribution patterns. And we are now trying to understand these less typical combinations. We have seen models, like Pat Wyman's, from which one might infer an hypothesis about why these combinations occur. And we have offered some hypotheses of our own. But all of these remain to date unconfirmed.

In this paper, we would like to put all of these models aside for the moment, and suggest another possible way of utilizing the statistical data that has been generated on the Enneagram and the MBTI to help individuals to come to a better understanding about what their MBTI scores might be able to tell them about their Enneagram type and vice versa.

Section Two: 'Microcosmicity'

skip to section three

What we want to introduce in this paper is a particular method for organizing and 'analysing' the patterns that occur in the individual's Enneagram scores, a very simple method. The method is one that is independent of any particular model for explaining the Enneagram-MBTI relationship. You don't have to accept the 'Fudjack/Dinkelaker' theory, or any other theory, about which MBTI prototypes fit which Enneagram points in order to use the method.

Take the individual's top four Enneagram scores from the RHETI, and, using the table generated from the EM survey, determine which MBTI type statistically has the highest distribution in those Enneazones.

Lets take, as real life examples, the two individuals who were the INFJ artists in our first issue of 'The Enneagram and the MBTI'. The first, the artist mentioned above, did the portrait called 'Mary', and had as his top four Enneagram scores 4, 1, 5, and 7 - in that order. When you look at the table, you discover that INFJs are most densely populated in enneazones 4, 1, 5 - in precisely that order. For INFJs as a group those are the ONLY zones with an I-value greater than '1.0' (ie, more densely populated by INFJs than would be expected if distribution were random).

The second artist, the 'logomaker', is also an INFJ, but one who turns out to be an Enneagram One. His top four enneagram scores are 1, 2, 5, and 4. Again, the 3 enneazones to which INFJs as a group gravitate are amongst his top four scores. But, in this case, they are appear in a slightly different order. In addition, in his case, zone Two also makes a significant appearance, and one that is, from the point of view of INFJ demographics, surprising. This may be why one observer mistakenly identified him as a 'Two', and also what threw the observer off the track of guessing that the artist was an INFJ.

These three zones that are statistically signicant for the group of all INFJs (4, 1, 5) appeared in the top four Enneagram scores of yet another INFJ to whom we recently administered the MBTI and the RHETI. The RHETI scores of this individual were - 1, 3, 4, and 5.

And, sure enough, we again see the same phenomenon occuring in other MBTI types. The MBTI and RHETI test scores of a number of INFPs who have recently consulted with us, have zones 9, 4, and 6 in their top four Enneagram scores, and these are the enneazones in which INFPs gather in highest concentrations. Those three zones are ranked slightly differently in the individual scores of different persons, just as with the INFJs we've described above. But they still fall into the top four enneagram scores for each person. Furthermore, just as in the case of the individual INFJs, for some of the INFPs, an additional Enneagram point will unexpectedly insert itself amongst the 9, 4, and 6 in the 'top four'.

Indeed, this phenomenon - the way that the individual's Enneagram scores seem to parallel the distribution patterns of the MBTI type to which the individual belongs - occured often enough in our experience that we decided to coin a term that we could use amongst each other to refer to it. When the top four enneagram scores of an individual with a particular MBTI type includes at least 3 of the enneazones in which that MBTI type as a group concentrates most heavily in, we began to talk about those scores as demonstrating 'microcosmicity'. We used this term because the pattern in the individual's score were a microcosm of the group's scores.

Does this phenomenon of 'microcosmicity' ALWAYS occur? No, we saw exceptions. Does it USUALLY occur? Often enough to warrant attention. But, unfortunately, we do not have at our disposal a large enough sample of individuals who have taken both the RHETI and the MBTI to warrant a conclusion. The EM survey did not collect MBTI and RHETI scores, and thus does not provide us with information at a level of specificity fine enough to make such a determination (ie, data collected through what we have elsewhere called an appropriate 'data lens'). Perhaps others have already collected data in this form. If not, it would seem to us that it OUGHT to be collected, and analyzed with respect to the phenomenon that we are calling 'microcosmicity'.

As I've mentioned above, we HAVE personally tested individuals whose scores do NOT demonstrate microcosmicity, at least in the strict sense as defined above. For instance, an INFP whose top RHETI scores are 5, 9, 7, and 4. Only TWO of the four zones that are most concentrated show up in this individual's top four. Or the individual who is an ENTP (ENTPs as a group are most densely concentrated in zones 3, 8, and 1) whose top FIVE scores are 7, 3, 8, 1. Again, not an exact hit, according to the strict definition of 'microcosmicity' we've established above. But enough to make one wonder if there is still not a rough approximation of this pattern in both the case of the ENTP and the INFP.

What would it MEAN if 'microcosmicity' (or some similar pattern) were shown to occur in individual's scores more frequently than would be randomly expected? To explore that interesting question more deeply we should first clearly recognize that one can easily imagine a whole spectrum of statistical 'samples' - ranging from samples in which microcosmicity occurs in ALL individuals in the sample group, to ones in which it is present in NONE of the individuals. In the former case, we might want to say that the individuals in the sample exhibit a 'fractal' relationship to the group as a whole. A fractal pattern, to put it simply, occurs when the same pattern displays itself at different levels. For instance, a letter 'A', that is made out of a lot of small 'A's, could be said to be fractal. Like this -

        A
       A A 
      A   A
     A A A A
    A       A
   A         A

Not all part-whole relationships are fractal, of course. The following one is distinctly NOT fractal -

        B
       D E 
      F   A
     G A B A
    D       D
   E         F

And some, may exhibit something that is not exactly fractal, but 'close to it'. It might be instructive, in trying to understand the significance of 'microcosmicity', to explore what it were mean if the Enneagram-MBTI data were fractal in the strictest sense. For sometimes the 'limiting' case tells us more clearly what might be going on, to a lesser extent, in less 'pure' instances.

If all INFJs enneagram scores were exactly the same as the first artist we discussed above - with 4, 1, and 5 as the first three highest scores (duplicating the 4, 1, 5 distribution of the group of INFJs), all INFJs would, obviously, score as 'Fours'. (I know that it is not always the HIGHEST score that indentifies the individual's type, but for the purposes of this discussion, let's simplify and assume that it is.) We don't imagine that this purely 'fractal' parallel happens all that frequently.

If, instead, the group of all INFJs were like the three INFJs described above, they might test as different enneagram types but nonetheless each demonstrate 'microcosmicity'. Such a situation would seem to suggest that the distribution patterns of the INFJ are not cumulative 'averages' of very diverse preferences, but instead reflect a set of preferences that tend to be inherent in each and every INFJ. Might this not be construed as at least prima facie evidence that the Enneagram Points are NOT independently defined factors?

Section Three: Three Groups

skip to conclusion

In our experience, there are three groups in which individuals can be classified with respect to the phenomenon of microcosmicity.

The first (which we will call GROUP A) is the group that the three INFJs described above fall into. Lets imagine these three individuals participating in a survey similar to the one EM conducted. They would have been asked their Enneagram type and their MBTI type, but no further specific information regarding these scores. According, a table like the following one might be used to summarize the information they'd submit -

GROUP A
combination	frequency
infj-4	1
infj-1	2

It would look like the INFJs in this tiny group are typically Enneagram Ones. But what happens if we were to include just a few more individuals - some INFPs - and also ask these individuals for details about their top four enneagram scores? The following table might result -

GROUP A
combination	frequency	top 4 enneagram scores	statistical distribution of relevant MBTI type
infj-4	1	4,1,5,7	4,1,5
infj-1	1	1,2,5,4	4,1,5
infj-1	1	1,3,5,4	4,1,5
infp-9	1	9,4,6,2	9,4,6
infp-4	1	4,9,2,6	9,4,6

And if we recognize in this pattern the phenomenon of 'microcosmicity', we begin to see that the top score ALONE (or even a determination of the individual's enneagram 'type') may not be the most important consideration in trying to understand the meaning of the distribution of MBTI types across the Enneagram. When we do not limit our investigation to the mere tip of the iceberg, statistically speaking (the top enneagram score of the individual), we are less likely to take the second and third individual in the above table as merely counter-examples to the proposition, based on wider statistical samples, that the infj is in some sense 'prototypical' of enneazone four.

Another way of saying this is that patterns such as the one in the table above will of necessity remain INVISIBLE insofar as studies do not accumulate data at the appropriate level of specificity to reveal such patterns. This is what Pat and I had in mind in earlier papers when we spoke about the use of different 'data lenses', at different levels of specificity. Walter Geldart has frequently made a similar point, on the basis of his 'enneagram of process' approach, and tends as a result to treat the enneagram PRIMARILY in terms of such PATTERNS in the enneagram scores, as opposed to focusing primarily on a specific identification with one enneagram point as a 'type'.

In addition to individuals who display strict 'microcosmicity', like the individuals in GROUP A above, there are also those who tend toward it, but are not 'pure' examples -

GROUP B
combination	frequency	top 4 enneagram scores	statistical distribution of relevant MBTI type
entp-7	1	7,3,1,5,8	3,8,1
infp-5	1	5,9,1,4,6	9,4,6

The second person in the above list is of particular interest. One might not expect an IF in a zone that is characterized by IT, but despite this atypical Enneagram-MBTI combo, this individual's top scores also displays preferences for all three zones characteristic of infp - 9, 4, and 6!

The third, and least populated group, includes be those individuals who seem not to exhibit 'microcosmicity' at all. This group, in our experience, seems to be the smallest of the three -

GROUP C
combination	frequency	top 4 enneagram scores	statistical distribution of relevant MBTI type
infj-2/9	1	9,2,3,6,7,8,4,6,1	4,1,5

This individual, who scores low in 4, 5, and 1, also does not relate to the verbal descriptions associated with these points. At first, this individual appears to CONTRADICT our expectations for the INFJ. In this particular case, however, there is some question as to whether this individual is actually an INFJ, as his F score is very high, and his functional preference order, as indicated by the FD33, is F-N-S-T, the order that is traditionally associated with the INFP, who concentrates most strongly in enneazone 9.

It is interesting that the individuals in GROUP A appear to be reasonably confident about their type identifications, individuals in GROUP B seem to express more confusion about their types, along with a tendency to bounce about in trying to come up with a satisfactory identification, and those in GROUP C seem to be least confident of all, with a tendency to 'try out' numerous type identifications, and remain generally dissatisfied.

Conclusion

skip to footnotes

My purpose in writing this paper was to describe some observations that we have made recently, regarding a phenomenon that we have called 'microcosmicity', and discuss a method for analysing Enneagram and MBTI data that might prove helpful in understanding the relationship between the two systems. We haven't collected sufficient data in a form that would be necessary to predict the degree to which 'microcosmicity' might be present in the scores of individuals in the population at large. But there are some general conclusions that might be drawn from the observations that we have made of individuals we have tested and talked to.

1. It is not necessary to attribute any special VALUE to belonging to one group rather than another. Individuals whose scores exhibit 'microcosmicity' don't necessarily seem to be 'more healthy' than those that don't. Nonetheless, having scores that are 'microcosmic' with respect to the MBTI group to which they belong may very well be less confusing or tension-producing to individuals, and hence more socially satisfying.

2. The pattern that we are calling 'microcosmicity' appears to be hard to discern in the absence of specific Enneagram scores. There may thus be a hidden advantage in the utilization of formally administered Enneagram tests, if 'microcosmicity' turns out to be a significant feature. This is worth mentioning, because many Enneagram practitioners - Pat Wyman and Michael Huber are two who come to mind - rely almost totally on typing individuals without using a formal instrument.

3. Methodologies that seek to understand the relationship between the Enneagram and the MBTI solely on the basis of the Enneagram and MBTI type identification, without considering patterns in each set of 'scores', may be taking too simple a view to be very useful in the long run.

In addition to the above conclusions, the methodology that I have presented here generates a number of general hypotheses that could easily be tested by merely collecting the MBTI and RHETI scores of a sufficient number of individuals -

1. The individual's Enneagram scores will in general exhibit 'microcosmicity' when compared to the distribution patterns that the MBTI group to which they belong.
2. People who have 'compatible' MBTI and Enneagram types will more frequently exhibit 'microcosmicity' in their scoring patterns than those who do not.

In the past few months we've had discussions with a number of individuals who are confused about their Enneagram type. In some instances it has proven helpful to use the method described in this paper to assist individuals to begin to organize their enneagram scores into patterns that are meaningful in MBTI terms, creating a bridge between the two systems. This is especially useful with individuals who, for whatever reason, relate more easily to the MBTI as a typology than to the Enneagram.

Although the method described here can be used in co-ordination with specific MBTI-Enneagram theories, it does not depend on any particular model for discerning meaningful Enneagram-MBTI relationships.