and Kinship in Rural Mexico - The
D. White, H. Nutini, L. Brudner and M. Schnegg
initial text of article
Some first impressions of the Compadrazgo Network
The colors mark different structural
positions in the
From this classification of people a classification
of links can be generated. There are for example links the well integrated
Belenos (yellow) have to the not integrated outside world (red). Those links
are likely to different from the links - let's say - the well integrated
Belenos have to other well integrated Belenos (links from yellow to yellow).
The general question is whether links that are structurally different differ
also in other matters.
In the classification we find seven different
For a first analysis those patterns will be combined to only three. These are the tree like ties that link to the outside world (3 and 6), those that link to the outside but integrated non Belen community (4 and 7) and links that stay within Belen ( 1, 2, and 5)
A first question would be through what kinds
of compadrazgo those structurally different links are established. Here is
a the raw data:
Correspondence Analysis helps to make some of the structural patterns visible. Since the matrix rank is 2 we have a perfect fit in two dimensions (60% variance on the first eigenvalue which means that you primarily need to look at that dimension).
Here are the results:
The red dots are
the sacramental compadrazgos. The analysis clearly shows that the most important
relationships are the tree like ends of the networks. These important
relationships are clearly not established within the community
nor within the redundant block that surrounds the community. The explanation
is pretty strait forward: the structurally weakest ties (the
dead ends) become the highest symbolic or cultural emphasis. Another argument
would be that people try to integrate far away but potentially important
people in their personal
In which regions do those structurally
different ties reach
There is a general tendency that most of the integrated ties stay closer (in physical terms). Some cases are interesting and ask for an ethnographic explanation.
1) Why is it that Acxotla del Rio and Los Reyes are so well integrated? Acxotla is a town on the highway from Belen that is the edge of the capital of Tlaxcala and is the "urban settlement" for Belenos; Los Reyes is nearby and a bit further from the city.
2) The tree like relationships go Tlaxcala, Puebla, Mexico D.F, Contla, Santa Ana Chiautempan, Apizago and San Matias. Tlaxcala and Mexico D.F. are the state and national capitals, Puebla the capital of the neighboring state, Contla is the regional weaving industry, Santa Ana and Apizaco the nearest market towns, and San Matias the next village along the highway in the direction of Santa Ana and Tlaxcala city.
3) What is the difference between San Bernabe
and Santa Ana Chiautempan? San Bernabe is a neighboring village similar
to Belen; Santa Ana the closest market town.
Ties and occupations
Do people with different occupations have different patterns of ties? And if yes - is there any change over time?. The following analysis shows how the proportion of ties that goes in the relinked block has developed over the last years.
The analysis shows that there was an overall
peak in the late 50s and 60s where many relationships were invested into
relinked ties. Thought that pattern is not the same for all occupations:
the labradores show a very stable pattern, some others (like the
oficios corrientes invest more ties in the relinked structure than
in tree like relationships to the outside. If we look at the picture only
from the latest point in time there are some differences in how much the
group spends in the community.
Social Class and Stratification
Which role does the economic rank of the two compadres play for the establishment of the relationship? In theory we can distinguish at least three different systems: a class system, where almost all relationships are within ones social and economic class, an egalitarian system, where the economic position is not reflected in the social relationships and a patronage system, where wealth differences can be expected to be very high. Let us take a look, which model fits best to the compadrazgo data the relationships within Belen? Here is the matrix of relationships:
To understand more about the relationship between
the four economic groups it helps to convert the matrix into a vector of
distances between the groups. The distance between the "very rich" and the
"very poor" can be defined as 1 - 4 = -3. The other way around the distance
would be 4 - 1 = 3. This way we keep the asymmetry of the
relationships. The results look
How to read the table? The zero in the middle are all the relationships that are along the diagonal. If we were looking at a class system most of the relationships should concentrate here. This is not the case. In an egalitarian system all relationships should be equally distributed. This is not the case as well. What we find is that most relationships are one step below or above the class one belongs to, and these relationships are the only ones with higher than expected frequencies. In addition, there are more relationships that go up the economic ladder than there are relationships that go down.
But who is it that establishes those relationships? Is there any pattern here? Let us move a step further and consider which of the relationships in the matrix above are stronger than would be expected by chance. This statistical logic might uncover some of the patterns in the matrix. Here are the results:
The diagram shows all the relationships from
the above matrix that are larger than would be expected from the row and
column distributions. The first number on the arrow gives the number of
relationships and the value in brackets indicates how much it is above the
expected value. The direction of the arrow indicates who asked
Only because there is not much relinking in the elite does that mean that the elite does not have a special structural position?
BetW (s.d.) cases
As it looks it does not. The above
table gives the mean results from betweenness centrality measures
for each of the four groups. The results clearly indicate that centrality
in the system declines with the economic rank of the actor. The elite does
not "close up" by high internal integration but becomes an integrated
center of the network. This might be a somewhat typical situation
for a community that is in a transition stage. No class but structural