The Manifesto | The Dominion | Texts and Articles | Review | Links | ACCUEIL (FR) | HOME (EN)


Cosmic Influences: A new Proposal
by Graham Douglas


Ed. N.: Graham Douglas has written several articles concerning structural and anthropological research in astrology. Some of them have been published in Kosmos (Los Angeles), Correlation (London) and Semiotica (Berlin/New York). A short version of this text will appear (or has appeared?) in the Letters section of Correlation (vol. 20(1), 2001).
 

Most of the recent discussions of the Gauquelins' research have centred around allegations of fraud, (Dean (2000), Ertel (2000, 2001), Ertel and Irving (1996)), and neglected the phenomena. Since the collection of new data has become more difficult due to new laws protecting privacy, it seems that active research has stalled. Fraud implies the attempt to make data fit an existing theory, but if new theoretical approaches which the Gauquelins did not consider could be brought to bear on their data the question of fraud would become irrelevant. I hope to show how this might be done, using a small set of data from several sources.

I refer to one aspect of the recent discussion between Geoffrey Dean and Suitbert Ertel on the possibility that parental data tampering or self-attribution could explain the Gauquelin data.

In Note 14 of his article Dean examines the relation between eminence and family traditions of work (Dean 2000: 48, Ertel 2000: 75). I would now like to look at this in more detail because a crucial point has been missed, which in my opinion could change the direction of debate and research on the Gauquelin data.

The main source I am using is the monumental research contained in the book Born to Rebel by Frank Sulloway, not mentioned by either Dean or Ertel. Sulloway makes a simple but important point about siblings, which serves to put the discussion of possible planetary effects more accurately in context. This is that on average siblings share only half their genes with each other, and their frequently observed differences in personality are attributable less to heredity than to the need to occupy different "niches" in the family system, (Sulloway (1996): 88-89, 351 ).

Gauquelin's whole theory of planetary influence on the other hand relies heavily on genetic factors (Gauquelin 1976: 172-186). If the heredity connection fails so does the whole theory, which leaves a problem in explaining the professional data. Referring to the fact that most eminent people come from the 5% of most wealthy and intellectual families, he continues: "The planets have no role in this social aspect of success. As far as the movements of the planets are concerned everyone is socially equal." (M.Gauquelin 1976: 60). On the next page he continues to develop the point saying that only a small proportion of these children actually achieve eminence, and that this depends on the second factor : character, which of course is the cue for the Character Traits Hypothesis, (M.Gauquelin 1976: 60-70, abbreviated to CTH in what follows). My contention now is that the conclusion that social factors are irrelevant is hasty, and a proper understanding of what might be called family ecology leads to a new hypothesis of the mediation of planetary influence: by inherited parental character and by the part of character developed in response to the dynamics of the family system. A sensitivity to planetary influence is of course required for both factors.

Before describing Sulloway's work it is useful to summarize some points that that any theory needs to explain.
 

1. Gender Effects

Gauquelin analysed the data by comparing strengths of Chi-squared values, and argued that this showed no gender effect on planetary heredity, as expected from genetic theory (Gauquelin, M. and F. 1972: 144 ). However, as Dean points out effect sizes are what matter in this type of comparison (Dean 2000: 33) and he thus found that gender does play a part, and Father-Son effects are stronger than the others, especially Father-Daughter.

Other relevant work on gender has been done by Ertel (1995,1996) and Ruis (1995). The first demonstrates that eminent women have consistently higher G% levels than men, (Ertel 1995 : 10-14), except for those at the very highest levels of eminence. As Ertel comments, " Any theory that might be suggested to explain the Gauquelin effect would have to also take account, from now on, of observed gender variation." (Ertel 1995: 13). In another article Ertel analyses Ruis's finding (Ruis 1995) of gender differences in the diurnal distributions of the Gauquelin planets for married couples. Ertel shows that the difference disappears when the males and females are paired off by birth date, (Ertel 1996:6) instead of by marriage. He thus concludes that there is no gender effect for ordinary people, but this does not dispose of the question: does such an effect exist for ordinary married couples ? In fact it is evidence that this may be the case, and this brings us back to the importance of the family as an interacting system.
 

2. Selective Pressure and Eminence

In Gauquelin's model inheritance of character involves inheritance of sensitivity to a particular planetary influence too (Gauquelin 1976: 211-213 ). This is necessary to avoid the unlikely alternative that planets somehow rearrange the DNA of the foetus.

Any theory involving genetic factors needs to explain what selective pressure gives them scope to act. Although Gauquelin does not discuss this, it may be presumed that the reason the effect is expressed only in eminent families is because it confers the advantage of allowing siblings to profit from the fertile ground created by an established tradition of social (and generally economic) success in a particular field, via CTH. Eminence, and hence greater reproductive fitness would continue into the next generation. However it could be argued that eminence in parents is only a general advantage which favours eminence for siblings no-matter what their character or chosen profession.

In non-eminent families different selective pressures can be imagined. Without a family tradition there would be no advantage to the foetus to time its birth to repeat the parents' planetary patterns, but a strong MA or JU emphasis would be expected to confer a general personality advantage in social competition.

In discussing genetic factors it is important to keep in mind that many generations are often required for characteristics to become heritable. This implies that due to inevitable social mobility over long time periods, any heritable sensitivity to planetary timing would become dispersed in the whole population. It was presumably for this reason that the Gauquelins decided to look for heredity effects in ordinary people. However it seems to have become generally accepted that the heredity effect is not real simply because the second study failed to replicate, even though more recent results show a positive correlation again (F.Gauquelin 1992).

These considerations all point to the importance of the social context of the family, which Dean and Ertel have referred to many times in other parts of their recent articles without addressing the question of selective pressure. I do not claim to provide an exhaustive interpretation, but I hope it is clear that these issues require discussion if a convincing theory is to be developed.
 

3. Ertel's Curvilinear Eminence Effect

This curious result receives no explanation at all so far, but we can now ask: is it part of the genetic factor or the social one ?
 

4. No Gauquelin Effect with the Sun or Outer Planets

There are, at best only weak correlations of character or profession with the Sun, Mercury, and the outer planets. There is no obvious explanation based on strength of planetary influence.
 

5. The Vexed question of CTH

Gauquelin considered CTH essential to his theory, although Ertel (1993: 8) has disagreed. Despite the evidence for a character typical of each profession, the validity of the Gauquelin work on trait words extracted from biographies is a contentious issue, with Ertel claiming that it is only the result of biased extraction methods (Ertel 1989,1993, Irving 1993, Gauquelin and Tracz 1991). However, as Douglas (1997a) points out, others have achieved objective trait extraction techniques using inter-correlations between different extractors as a filter to decide which traits are truly representative (Simonton 1986, 1988, Gough and Heilbrun 1965). In this way Simonton was able to factor-analyze the traits extracted from the biographies of US presidents. Surprisingly, this point has not been taken up so far in the debate on CTH.
 

A New Approach

In considering these five points it is important to remember that there are two contributing factors to eminence: one inheritance of character, and the other social environment. In Gauquelin's analysis he proposes that the planetary effect is mediated by the first, simply because it seems absurd to believe that social class depends on birth time. I believe that environment is crucial, but in order to understand how this might be it is necessary to look at one particular social environment: the family and its interpersonal dynamics, as described by Sulloway. It will then be apparent how most of the above points can potentially (the relevant data needs to be collected ) be accounted for.

Essentially the discussion which follows suggests a shift of attention away from genetics and inheritance of individual personality traits, towards interpretations which give more weight to inter-personal dynamics, and the timing of births. In this view there must still be some kind of planetary sensitivity, but instead of being transmitted only by inheritance of parental character it is suggested that each foetus "knows" its eventual position in family birth order and its birth is timed ( in terms of G+ zones) to promote the personality attributes most adapted to that position in the family.
 

Evolutionary factors

The link to character occurs in 3 steps:

(1) Sulloway begins from a recent extension of Darwinian theory known as kin selection. This works through the principle of inclusive fitness, by which it is families not individuals who need to survive, so inclusive fitness measures the contribution an individual makes to their own and their kin's reproductive fitness, (Sulloway 1996: 58-60).

(2) One of the major consequences of this is conflict between parents and offspring over how much attention and resources are invested in each one, but in its application to humans the drive to maximise inclusive fitness has to be filtered through the cultural context (ibid.:64). Thus in societies where chance death or loss of fortune occur frequently, equality of parental investment in children is common, as occurred in mediaeval Italian city-states, but in stable agrarian societies male primogeniture is the rule. In the latter system males are given an advantage because their reproductive capacity lasts longer, and wealth can magnify this capacity further because it allows a man to remarry if his wife dies. But in non-élite families sons are at a disadvantage compared to daughters because they cannot compete against wealthier men, while a sister can always improve her status by "marrying upwards" ; in such cases parents invest in girls through the dowry system, (ibid. 66). This type of behaviour also of course works as a link between eminence and inclusive fitness. There is not space here to go into further detail but it is time to note two new factors which have come into the equation: gender and birth order.

(3) The next step in the chain of argument concerns the adaptive strategies of siblings living in a situation of conflict over parental investment. By considering these in terms of the five major personality dimensions Sulloway arrives at the following contrasting profiles, which are amply confirmed by published research (ibid.68-75):

Firstborn children tend to be more assertive, confident, dominant, antagonistic, unsociable, conscientious, accepting of authority, and conservative, while:

Laterborns tend to be less confident and assertive, more agreeable, sociable, rebellious against authority, unconventional and open to experience.

Only children occupy an overlapping position.

To give some idea of the strength of these effects, Sulloway found that professional revolutionaries were 18 times more likely to be laterborns than firstborns, (Sulloway 1996: 297).

It is interesting now to see what this might predict in terms of planetary personalities. It seems clear that firstborns will be stronger on Saturn (conscientious and conservative), but also strong on Mars (assertive, dominant, antagonistic) and Jupiter (confident). Laterborns will be weaker on Saturn (sociable, rebellious, unconventional) but also show significant Mars traits of rebelliousness, and Jupiterian sociability, unconventionality and openness to experience. Finally they will be stronger on Moon and Venus traits of sociability and agreeableness, and again unconventionality. So , while the predictions are mixed for Mars and Jupiter, they seem clearly dichotomized for the other Gauquelin planets: Firstborns will show a stronger SA, and a weaker MO and VE influence, and laterborns will show the reverse pattern.
 

Interaction with Gender

Sulloway distinguishes Interaction Effects from Main Effects. A Main effect occurs in the same way irrespective of the behaviour of another variable, although its magnitude can be combined additively with other effects. An Interaction Effect, on the other hand, is one which has non-additive or emergent properties in conjunction with another variable. Two of the most important interaction effects on birth order occur with gender and social class. As an example we can take conformity, which as a main effect decreases from firstborns to laterborns. However there is an interaction with gender so that while this pattern holds for pairs of sisters who also have brothers, it is reversed in all-girl sibships (Sulloway 1996: 150).
 

Relevance to Astro-Research

The most obvious outcome of the discussion so far is the clear prediction of differences in G-sector positions for SA, MO and VE between first- and laterborns. This prediction depends on CTH but takes no account of profession, which seems to be a separate factor independent of inclusive fitness. Although as noted above, eminence does potentially correlate with inclusive fitness.

Two alternative hypotheses suggest themselves, not necessarily mutually exclusive:

1) Planetary strength is related mainly to birth order , leading to a tendency for SA professions (scientists and physicians) to favour firstborns, and MO, VE professions, (writers, musicians and painters) to be pursued more by laterborns, or

2) Planetary strength is related mainly to parental profession, and the ecology into which children are born is sufficiently "coloured" by it to produce a tendency in all the children to be born with the planetary pictures of one or both parents.

As they stand neither of these possibilities is adequate to explain the Gauquelin data, because MA and JU do not have a clear place in the first alternative; and because in the second, it is not clear how factors encouraging children to follow their parents' professions automatically increase family success beyond that resulting from the privilege of parental eminence. This clearly depends on the social conditions of the time, and in many cases a change of profession could bring greater success. These questions cannot be decided without collecting the relevant demographic data. Because this approach is new however many conjectures can be considered.

Another relevant observation in Sulloway's book emerges in his analysis of the personae on the two sides of the Reign of Terror in 1793. While the same major contrast between first- and laterborns holds as before, the contrast is not only between those laterborns who were liberal and rebellious and those firstborns who were conservative and centralist, but also involves the extreme violence practiced by those trying to hold on to power. Sulloway points out that a full picture requires two dimensions of political behaviour to be recognized, the contrast of Radical-Conservative and the contrast of Tough-Tender mindedness (Sulloway 1996: 286). When this is done it becomes clear that firstborns and laterborns are in opposite corners of the square. Firstborns tend to be tough-minded conservatives (fascists and absolute monarchists), while laterborns tend to be tender-minded liberals. The progress of the French Revolution can be followed in terms of the percentage of firstborns in power each year from 1789-1794. The results are quite extraordinary, showing a fall from 50% to 20% until the beginning of the Reign of Terror, when it rapidly increased to 70%. In terms of the two dimensions there had been a shift from tough-minded conservatism to tender-minded liberalism represented by the Girondins and the Plain groups, and then a move to tough-minded liberalism as the Montagnards took control. In this trend there is another interesting observation, which is that in the most tender-minded period there was a majority not just of laterborns but specifically middleborn people, a trend exactly paralleled among the main figures in the struggle for black emancipation in the USA, (Sulloway 1996: 300).

These considerations suggest a little more detail in assigning possible planetary patterns. First it seems clear that political fanaticism is linked to firstborn children, but it also has a correlation with a high G% for JU, (Gauquelin 1976: 157-159), which is true of politicians in general. There may also be some overlap with both JU and SA traits, via MA and among the most ruthless people it seems likely that a strong MA would combine with the toughest traits of JU and SA, (Gauquelin 1976: 146-147, Douglas 1997b). Next the tender-minded liberalism of middleborns seems quite typical of MO and VE character traits (F.Gauquelin: 1982).

Putting these observations together we can place the planets on the map of political attitudes (which also holds for attitudes in general (Eysenck and Wilson 1978).



Attitudes, Planets and Birth Order


At this stage it has to be recognized that these assignments are only suggestions, and each planetary zone in the diagram may be spread out and overlap its neighbours. However the pattern in Fig.1 does broadly preserve the necessary semantic oppositions which require MA to be far from MO and VE, and JU to be far from SA (Douglas 1995, Müller 1992).

Now it is time to consider how a theory based on birth order deals with the 5 questions noted earlier.
 

1. Gender Effects Revisited

Sulloway's work suggests that gender-specific effects are highly likely, since they clearly influence the way firstborn children may be less dominant than laterborns in all-girl sibships, (Sulloway 1996: 150). Secondly if we consider the more general concept of the family "ecology" then the existence of gender differences such as those found by Ruis (1995) between marriage partners may also "colour" the dynamics of the family system. In contrast, Gauquelin's hypothesis of parental character inheritance does not predict gender variations, as already mentioned.
 

2. Selective Pressure and Why only Eminent Families?

Unlike the model of purely individual genetic transmission of planetary effects, a family ecological approach suggests a plausible role for the planets in increasing the viability of the family system. Thus if the timing of births (both on a scale of years and hours) occurs according to a rhythm synchronized with the planets in some way the result could be that children are born with personalities best adapted to the particular niche which birth order creates for them. This would lead to less destructive sibling and parental conflicts, and therefore more functional families. A more functional family would in turn have greater survival ability, and hence better reproductive fitness. (An immediate corollary would be that dysfunctional families would be an interesting source of new astrological data, focusing on the interactions between siblings.)

If there is a planetary sensitivity which can cause the foetus to time its delivery to correlate with a certain plus-zone pattern then it seems clear that this is likely to become connected with birth order too, through the connection between family functionality and inclusive fitness. And if this is a selective pressure it must have existed since before man emerged from the apes as a separate species.

It is now possible to suggest a more specific version of Gauquelin's point about social class. While agreeing that any planetary influences do not distinguish between social classes, once they are established, we can also allow the possibility that the sensitivity to planetary influence began to have an impact because functionality deriving from a stable role structure led to greater inclusive fitness, and in turn greater success in social competition. This would then have fed back offering lower levels of infant mortality, and greater health for successful families. The implications for family size are also important and have clearly changed over the last century for other social reasons.

Thus a theory relating planetary effects to inclusive fitness suggests how they may help families achieve and maintain eminence. In contrast the theory of genetic transmission offers only a general link by which family eminence will be maintained if children have the same G+ planet as one of their parents. As mentioned above however, eminence and social status help to open all professional doors, not just those of the parents profession.

Considering the specificity of planetary effects another interesting difference emerges between the parental inheritance and the family ecology theories. Thus while both are quite compatible with inclusive fitness, only the latter requires a differentiation of roles and personalities within the family. According to the former theory the best result would be for all the children to acquire the same plus-zone planets as one of the parents.

The ecological model used here is based on roles, but another interpretation is suggested by a recent analysis of the effects of competition on Circadian rhythms, (Daido, cited in The Economist 2001). Noting that many such rhythms vary considerably from 24 hours he suggests that under conditions of competition there is an advantage to all species in timing their activities differently to avoid creating a 'rush-hour', (Daido 2001). Apparently examples are known in which genetic mutations determine alterations in the period of Circadian rhythms, (Dunlap (1993), cited in Daido (2001).

Although this model was devised in terms of competing species there seems no reason why similar considerations should not be applied to the functionality of families, and this could be mediated via planetary timing of Circadian rhythms, which is a feature of Seymour's model of astrology, (Seymour 1988: 89-90, 107-108). Genetic links may play a part, but all that is necessary is some kind of sensitivity to planetary cycles, which can act as a timer.

Other aspects of evolution and selective advantage are mentioned in relation to planetary symbolism and colour perception (Douglas 1999).
 

3. Curvilinear Eminence Effects

Ertel has found that some eminence effects differ from Michel Gauquelin's proposed direct correlation between planetary strength (measured by G%) and eminence rank. He found this effect with several palanets including SA, the planet most closely associated with eminence in science (Ertel 1993b). In Sulloway's book scientists provide an interesting example of the complexity of interactions in this field. The raw data show a modest eminence effect preferring firstborn children, but when they are controlled for 2 other variables: social class and family size the effect disappears, (Sulloway 1996:109-110). Firstborns tend to enter science more frequently, largely because they are economically favoured by parents and better educated (Sulloway 1996: 254). However Sulloway also mentions that while firstborns are not relatively more frequent among eminent scientists (because of the relation just described), they are more common among all scientists including non-eminent ones, (ibid.: 482).

So if the planetary hypothesis constructed above is correct there will be a preponderance of scientists as a whole with SA in plus-zones, but the correlation with eminence will become weaker as eminence increases, exactly as Ertel has found (Ertel 1993b). It is also worth saying that the criteria for achieved eminence can change especially during scientific revolutions. Generally the tendency is for those who achieve eminence in the form of presidencies of committees, and medals to be conservative first-borns, but there was a period after 1859 when this changed: the Darwinian revolution began to make progress and the old guard of firstborns was replaced by laterborns.

In line with the general tendency for laterborns to be more open to experience and less conservative or authoritarian, they tend to achieve eminence by more unconventional paths. During the nineteenth century laterborns were more successful especially in biology because their ability to diversify was an asset over the typical firstborn tendency towards specialism. However, firstborns had an advantage in the lower classes because their class position also correlates with ability to diversify and this was a help particularly in biology, while physical science was closed to them because of the financial burden it imposed. Sulloway makes the point in summary that "paradoxically social class expresses itself as within-family differences", (ibid.:110).

The practical investigation of these effects may often require extensive controls for gender, family size and social class, among other variables. Although middle children are intermediate on most variables, different patterns have been found between these groups in some cases, so for example on militancy, political leaders show the lowest scores among middle-born children, (ibid.:300). It could well turn out that when enough data are available to be split into 3 groups new features will emerge.
 

4. Only Five Gauquelin Planets

It has been generally assumed that this observation requires an explanation in terms of the mechanism of planetary influence, so it is interesting to see what a family-ecological model has to say. The first relevant point, noted above is that this hypothesis works through role differentiation within the family system. So if we assume some form of CTH to be operating, the question of role differentiation requires changes in the diurnal placing of the planets, when the interval between successive births will be, say 1-5 years. The fact that the 3 outer planets do not move much in longitude during this time does not stop them occupying all possible diurnal placings, but it does mean that they will "move" as a group, so that the house placing of one will determine that of the other two. Depending on their angular separations this may mean that angularity in one requires angularity for another (in conjunction and opposition especially), or the converse. This issue is relevant to the angular separation of JU and SA also, which will change over a period of 2-3 years, (see Douglas 1984). With the Sun and Mercury, a different issue arises, which is that the strong bias of the nycthemeral curve will limit the full expression of differences attributable to diurnal placing.

The ecological model cannot do away with the need for an explanation of planetary influence.
 

5. CTH and Implications for Future Research

The question of CTH can be approached again in a new way through studies of birth order, because of its established correlation with the main dimensions of personality described above.

This theory also offers a new perspective on the observation that Gauquelin effects have weakened since 1950. The use of induced delivery obviously affects this, but the ecological model suggests that family planning will be expected to have an influence too, through altering family size and intervals between births.

In assessing the usefulness of this theory it needs to be born in mind that the Gauquelin data was not controlled for social class. This is not a weakness however if planetary effects are real, because they should move in parallel to whatever birth-order correlations exist for uncontrolled samples. With this in mind I will now present a small study which confirms the inferences drawn above on variation of planetary G% with birth order.
 

Results of a Pilot Study

As a preliminary test a small sample of timed births was obtained for siblings whose birth order was known. They are a rather heterogeneous collection of "VIPS", including european royalty, children of mainly Italian and US film stars, supplied by Grazia Bordoni; Scottish MP's and their siblings obtained from Caroline Gerrard; and the family data of Marian Bollen.

The planets were grouped into 3 sets as shown below. The first column is approximate primary plus-zones, here calculated by counting the planets in Placidus houses 9 and 12 together with one third of the adjacent angular house by longitude. Next are the secondary plus-zones calculated similarly from houses 3 and 6, (Ertel 2001: 45 for definitions). The advantage of using the extended sectors is an increase in number of subjects in the first 2 columns for a small sample size. The expectation values given in the tables are based on simple calculations of fractions 2/9, 2/9 and 5/9, which are not accurate because of differences in the times represented by opposing pairs of sectors above and below the horizon. Taking all four plus-zones together avoids this difficulty but would also have masked the most important patterns. These astronomical factors have no effect on MO, and practically none on JU, according to the table given by the Gauquelins (1973:44). All other expectation values have been corrected.
[The expectation values used in Chi-squared calculations are unchanged for MO and JU. For MA and VE the expectation values for AS+MC sectors are multiplied by 1.03, and the DS+IC sectors by 0.97. For SA the pattern is reversed and the factors are 0.94 and 1.06 respectively. (M. and F. Gauquelin (1973): 44).]
 
 

Table 1: Combined VIPS, Royals and 14 Scientists and Philosophers


198 Firstborns

AS + MC

DS + IC

Rest

Total

SA *

57

32

109

198

JU

44

54

100

198

MA

54

38

106

198

VE

52

42

104

198

MO

44

45

109

198

Totals

251

211

528

990

Expected

44

44

110

 

325 Laterborns

       

SA *

70

50

205

325

JU

84

68

173

325

MA

73

71

181

325

VE

81

54

190

325

MO *

92

60

173

325

Totals

400

303

922

1625

Expected

72.2

72.2

180.6

 

 
 

Table 2: 155 Ordinary People


49 Firstborns

AS +MC

DS +IC

Rest

Total

SA *

18

6

25

49

JU

11

17

21

49

MA

10

6

33

49

VE

13

13

23

49

MO

13

11

25

49

Totals

65

53

127

245

Expected

10.9

10.9

27.2

 

106 Laterborns

       

SA

15

26

65

106

JU

30

27

49

106

MA

34

23

49

106

VE *

30

14

62

106

MO

21

33

52

106

Totals

130

121

274

530

Expected

23.6

23.6

58.8

 

 

Significant Deviations

Table 1

The first feature which is evident in Table 1 is that when all the plus-sectors are combined there are few deviations from expectations. A small and non-significant excess of JU in both first and laterborns is apparent together with a small deficit mainly in laterborns for SA. This picture is what would be expected with a sample including a fraction of politicians and actors, but it is interesting that the SA deviation is mostly accounted for by laterborns only.

The most interesting patterns are seen in comparisons between the primary and lesser plus-zones, where there is a proportionately larger excess of primary sector SA in the firstborns compared to laterborns, and both deviations are significant. While the deviation in the laterborns mainly consists of a low frequency in the secondary zones offset by an excess in non plus-sectors, in the firstborns the excess in primary plus-zones correlates with a deficit in the secondary sectors alone. Still, taken together this amounts to a clear confirmation of the hypothesis of a stronger SA effect with firstborns, and for this reason the probabilities (which relate to a test of the null hypothesis) should be divided again by 2, to get a one-tailed significance.

For SA, Chi-squared with df = 2: SA(Firstborns) = 10.46 (p = 0.005); SA(Laterborns) = 12.50 (p = 0.002).

Turning to the other prediction for MO and VE we see strong excesses among the laterborns and no significant deviation in the firstborns, as measured by the relative frequencies in primary and secondary sectors. Chi2 for MO(Laterborns) is 7.80 (p = 0.020). Although VE is not significant the combined data for VE and MO yields a value of Chi-squared (df = 2) of 6.8 (p = 0.034).

There are no significant deviations for the combined 4 plus-sectors, as can be seen from the third column which is almost identical to the expected frequencies, except for SA(Laterborns) where Chi2 (df = 1) is 7.4, p = 0.007. However the positive deviations in the key sectors (first column) are the usual measure of the Gauquelin effect, and when this is tested against all other sectors combined, a significant result is found for MO, with Chi2 of 6.98, p = 0.008.

The predictions for MA and JU were less clear, because traits were identified on both sides of the firstborn - laterborn character contrast, so it is interesting that the deviations observed are smaller. However there are still interesting patterns which need to be explored further, suggesting for example that the dominance component of MA is stronger than the rebellious one, and that the openness to experience and sociability of JU are more important than its dominance. Of particular interest is the pattern in which a deficit in the primary zones is offset by an excess in the corresponding secondary zone sufficient to give an overall excess in the plus-zones, (JU for firstborns).
 

Table 2

This is a collection of ordinary people consisting of 22 from the family tree of Grazia Bordoni, and 42 kindly contributed by Marian Bollen, from a study of 5 generations of a farming family in Holland (Bollen 1982). The original publication of the Bollen data concentrated on the inheritance of plus-zone JU, so it is not surprising that there is a small overall surplus of JU in Table 2, but no significant deviations for any other planet when the 4 plus-zones are taken together.

When the plus-zones are separated into the 2 groups in the tables the pattern for SA of a relative excess in AS+MC for firstborns is repeated together with its opposite in laterborns, (Chi2 for firstborns (df = 2) is 8.9, p = 0.012). In the laterborns there is again an excess of VE in the AS+MC and a deficit in DS+IC, (Chi2 (df = 2) is 5.0; p > 0.05), but the corresponding excess for MO is in the DS + IC. The result for SA thus replicates significantly in the second sample, while the VE result is in the same direction as in Table 1, but MO is not.
 

The Character of the Combined Samples

The data in Tables 1 and 2 were combined (making 247 firstborns and 431 laterborns), with first and laterborns added together, and all 4 plus-zones combined. When these are compared with the non-plus zones, the only effect which is statistically significant is still an excess of JU in plus-zones (Chi-squared with df =1, is 6.74 for JU (p = 0.009); 4.48 for SA, (p = 0.035), as would be expected since two of the samples show this effect. In other respects the sample seems to resemble a normal non-eminent population.

It is interesting now to examine how deviations vary depending on different ways of breaking the sample down into categories.

When the same test is repeated for the firstborns and laterborns separately it becomes apparent that most of the deviation for JU is concentrated among firstborns (Chi-squared with df = 1 is 4.30, p = 0.038), while it is insignificant for laterborns. There is now a significant deviation for SA too, but this time all concentrated among laterborns (Chi-squared, with df = 1 is 8.80, p = 0.003 representing the whole deficit of SA in combined plus-zones. Since there is practically no deviation from expected values for SA in the firstborns this test shows that the 2 samples are significantly different from each other in SA distribution with this probability.

Since the Gauquelin effect is expected to be strongest in the key-sectors (AS + MC in the tables), it is useful to consider these figures in comparison to all the other diurnal positions combined. When this was done the values of Chi2 (df = 1) for SA were strikingly different between first and laterborns: 9.46 (p = 0.002) and 0.36 (not signif.).

The combined data was then broken down by separating the primary and lesser plus- zones, while keeping the first and laterborns together.The values of Chi-squared, (df = 2) were calculated using the corrected expectation values.

The only significant result was SA, 17.43 (p = 0.0002). Most of this is contributed by the deficit in the lesser plus-zones, which is more than double the excess in the plus zones.

This last calculation was then repeated after separating the first- and laterborns, with the following results for Chi-squared (df = 2):Firstborns, SA = 22.20, p = 0.000015; JU = 6.63, p = 0.037. Laterborns, SA = 10.6, p = 0.005; VE = 8.87, p = 0.012; MO = 4.03, p > 0.05. MO has lost significance, but when MO and VE are combined the result is 8.55, p = 0.014.

As a final test 2 sets of expectation values, for the first and lastborns respectively, were derived from the weighted mean frequencies of each planet in the 3 columns in the combined sample. This procedure eliminates all significant deviations (df = 5) in both first and laterborn subsamples. However when the lesser plus-zones are combined with non-plus-zones and weighted means used, SA: Chi2 (df =3) is 9.85, p = 0.019. This probability is the level of significance at which the first and laterborns are different from each other, giving evidence that the effects are real.

However it does not answer the question of why both subsamples and the combined set show such large deviations from the normal expectation values for SA and MO+VE, especially in the lesser plus-zones, and why they show no significant result for MA in either sample. While the JU result is predicted by the sample the SA deviations are far more significant and have no obvious explanation, except the theory presented here.

Large Families

Marian Bollen has also supplied me with another set of data from more distant branches of her family, and although a small sample this shows a very different picture for SA. On closer examination it was found that the new data was composed maily of families of 7 or more siblings compared to a range of 2-4 for most of the families in Tables 1 and 2. It will be particularly interesting to gather more data for large families, especially as there is a suggestion that they tend to segment into smaller groups, Hoopes and Harper 1987: 29), or that they require more cooperation, (Sulloway 1986: 99). These possibilities might suggest respectively that SA would be distributed at every 4th or 5th position, and that the firstborn would have a tendency to more VE or MO character to act as a force for cohesion.

Only Children

In this investigation only children were included in the data as firstborns. However since they are described by Sulloway as having some characteristics in common with both first- and later-borns it was decided to remove them and repeat the analysis. When this was done (removing 21 births) there was little change in the results, although the Chi-Squared values dropped due to the smaller numbers. The births removed often involved families of more than one in which there were large gaps between births, or it was known that the fathers were different, so that they were previously treated as two or more separate families.

Twins

The case of twins having different personalities despite being born at the same time is one of the oldest objections to astrology. Any theory which interprets birth charts by making a simple 1:1 mapping onto a set of meanings will face this problem, and the usual response of putting a heavy emphasis on very small differences in timing between the births seems to stretch credibility. The ecological model outlined here starts off by situating all the family members in the context of a field of mutual interactions, thus making clear that astrological influences are not the only ones. Given that such systems are frequently non-linear one phenomenon to be expected is divergence, in which very small alterations in environmental parameters can cause the system or its parts to move into one of two opposite modes of behaviour, (Woodcock and Davis 1978: 56-57). Even more interesting is the observation that discontinuous switching (known as a phase transition) occurs between two behaviours in such systems. This is also observed in cyclic behaviours which can switch from syncopated to synchronous rhythms at critical values of system parameters, (Kelso 1997: 46-53, and 93-95). I would speculate that twins would tend to opposite behaviours in small families to increase role differentiation, and towards similarity in large families. To put this in the psychological language familiar to many astrologers it could be said that the shadow-side of a character, or the other side of its gestalt, is also expressed in twins. Using astrological language an example would be opposite signs, so that a planet in Virgo expresses Virgoan characteristics but contains implicitly the characteristics of Pisces and vice versa. However under conditions of stress an inversion may occur allowing the implicit to take over.

A More Detailed Picture

Although the data reported in the tables is grouped into only 3 sets for statistical analysis graphs were also prepared showing the frequencies in each of the 12 Placidus Houses. Although the numbers are too small to say much some interesting features are evident. Thus, both Moon and Venus have strong fundamental (= first harmonics) waves for laterborns peaking in the AS region and at a minimum in the DS. Mars shows a strong 2nd harmonic wave for firstborns, and Saturn probably has significant 3rd and 4th harmonic components in both first and laterborns with different phases.

Perhaps these variations of the distributions of the planets can be seen as a reflection of the postulated differences in circadian rhythms. Although the deviations in planetary days from 24 hours are quite small except for the Moon, there are significant variations in the ratios of time each planet spends above and below the horizon, as shown by the correction factors applied to expectation values.

Final Comments

The combined sample when analysed in terms of combined plus-zones showed only one significant deviation JU which could reasonably be expected from the nature of the sample. Yet large deviations appear when plus-zones are split, and for SA these increase in significance as the samples are combined, and show significant differences between the first and laterborns. For all these reasons it is hard to reject the hypothesis that there is something special about the whole sample and about the first and laterborn components, which cannot be explained by astronomical factors, but which is predicted by the hypothesis derived from Sulloway's work on inclusive fitness.

An attempt at replication is obviously called for before it is clear whether this is just a peculiar artefact or a new and significant avenue for astrological research. If the latter is the case it could provide a new direction to the longstanding debate over the status of the Gauquelin data.
 

APPEAL FOR DATA
Anyone with data to contribute, no matter how small and especially of ordinary people, is invited to contact the author who is currently working on a follow-up study.
 
 

    REFERENCES


To cite this page:
Graham Douglas: Cosmic Influences. A new Proposal
http://cura.free.fr/xv/14doug1.html
-----------------------
All rights reserved © 2001 Graham Douglas

HOME
ACCUEIL
C.U.R.A.
PORTADA
Centre Universitaire de Recherche en Astrologie
Web site Designer & Editor: Patrice Guinard
© 1999-2001 Dr. Patrice Guinard