Impact of Social Networks on Conjugal Roles: Analysis of Role-Relationships, Summaries of Literature

Download Impact of Social Networks on Conjugal Roles: Analysis of Role-Relationships and more Summaries Literature in PDF only on Docsity!

p a ..

Conjugal Roles in Urban Environments ,

f l l ... 1 ~\ -) ,

~------------.---------~. -~ ~------

f;~~.

Li

~:" , t

1

J

r. Conjugal Roles in Urban Environments: A Selected Comparison

by Elaine Le\·/is

A thesis submitted to the faculty of Gradu'ate Studies in partial 'fulfillment of the requirements for the degree of ~laster of Arts.^ " + IF.,'

Departm~nt of Anthropology McGi1l University t40ntreal, Quebec Canada

August, 1978

, '

, 1

\ \

o

..

, 1

l '

RESUME', o Cette thèse "e'st un examen ,critique dl une branc-he particul ière de 1a sociologie de la fam'illé^ " où domine le travail d',Elizabeth Bott (1971) sur les relations conjugales A rôles partagés et à rôles dichotomis~s. Ces relations et leur articulation au système social sont analysées A la lumière de trois hypothèses: 1) la densit~ du réseau social affecte la fonne de~ rôles dans la relation conjugale; 2) les réseaus sociavx oD 1es sexes sont di chotom; sés produi sent des rôl es co~j uguaux dichotomi sés;

3. la prépondérance de la parenté dans le réseau soc.;a'l engendre des rênes .conjuguaux di chotomi sés. Ces trois hypothèses ont été evaluées 'par rapport auxj travaux publiés. Il a été découvert que 1a prépondérance l de' i a ~arenté' dans 1e (^) 'réseau' impliquait une dièho.tomie des sexes dans le réseau,. qui açcompagnait une : " y. dichotomie sexuelle dans le foyer./ Il a été suggéré que la nature de la corrununauté , au sein J de laquelle ces ' rôles pourraient être trouvés est un aspechcrucial. L~s rÔles conjùguaux dichotomisés ont été découvert exclusivement dans la classe ouvrière ou dans les corrmunautés rurales; H, les rôles étaient gén~ralement du type de "basell^ de Banton. Le caractère essentiel de cel1es-c'i est qu'elles sont des cO)l1TlLnautés fermées. [nf.in, en prenant une perspective théorique plus large, on a ,défendu ... la thèse que dans les communautés fermées. les foyers ~ont moins bien intégrés dans la-sàciet~ prise en"son entier'. L'esse'ncefSe la vie sociale ~e)prés~nte en tenne des rappor.ts directs entre groupes él émenta ires, et, pour l~s modalités

de l'organisation, ' socia.lé, il est "'ait appe1'au système de p~renté~' Ôn

a dMendu l'idée " qu,lil est nécessaire de faire porter l'analyse su~' la question de la situation du foyer dans sa communauté et dans la societé 'cela dans' le cadre d'~n mod~le de l'intégration sociale, plus généra~. , 1

-- '

\ (^) .. /:~ iTraduct;o~ par;,( D. Legros \ .. '" (^) " 1 Il^ "'

"

(

/

.'

; AC~NOWLEDGEt1ENTS 'Ill 1 wish to thank my thes;s 1 sapervisor, Dr. P.C.W. Gutkind, for his help and encouragement\d~ring the preparation and writing of \ ,

thi~ thesis. Thanks are also extended to Ors. M. Carroll an~

G. tinel ch, both members^.^ of my. thesi s committee, fot their expert assistance with this work. 1 also wi~h to aCknoWleël~e the help of Gareth Lewis, whose constant encouragement and w; 1'1 ingness to ta 1k at 1 ength about .conjugal roles resulted in the\comilletion \ of this thesis. Thanks are extended.to Mr. O. Legros for translating the a~stract' of this thesîs.

}

,, \ ~ ) \ (^1) \ cf^0 .•..,.,. ., [

.~

~-------------~---~- ~-----

v

,. (^) Chapter l -- IHE PROBLEM

f 1"'" t ...

'-, ) -,

.\

, -

/ -'^ ,

('

1 NT RODUCT ION This study , will con cern. itself with one particular aspect of family structure. Fami ly and kinship have al ways been a major area of anthropological research, largely becau~e kinship is a basic form of social organisation in pre-indJ,lstrial society. In this thesis, however, l shall be concerned exclusively with families in Great Britain and the United States, industrialized areas with less reliance on kinship as a mode of social organisation. Yet even in the industrialized setting the family is an important area of social research because. it is, at the very least, the unit of social ization, He unit oft residence and the unit of 1 ,consumption. _ Hence there are a v~riety of studies of, th'e fami,ly in industriaLized setting, but in spite of this research there remain +oblems ~s yet unanswered, r"ost importatltly, there is no cl ear understànding of what constitutes a family (see Rosenberg and Anspach, 1973:6-10) and disa'greement over^ •^ what constitutes^1 relevant data, In this study l shall deal with conjugal r~le-relationships. This is an excellent topic because there has been much research ,into it, jet there also remains much controversy ov-er .the theoretical implications of the differences in the fôrm of conjugal roles. The conjugal role-relatitonships i5 dominant and influential in any family, and through an understand1ng of this a better understanding of the family itself can be derived. Much as an understanding of internal dynamic's in the form of conjugal roles is important for understanding the .. family, so tao is

~., (

.

variation, and then, fram this understanding of the basic elements, ~ derive generalizatians which may, extend tleyond to an understanding of the fami ly as a genera1 construct. This is the intention of my study. 1 will start with a'simple aspect of family life, the conjugal 'ro1e-relationshi,p, describe its structure and attempt ta offer explanations for its variation based Jn a holistic scope',.

THE BOn HYPOTHESIS Elizabeth Bott is a good starting point, for she was th6! first to analyse tJ the role playeèl by the external environment ïn the formation of conjugal roles. In he}" book Fanlily and Social Network (1971) she investigated 20 famj1 ies, the conjugal role-relationship of. which ,var;ed- , from a joint or predominanUy sharing. to segregated, or strict division af labour" and activities. She attempt.ed to explain these difference~ th~ough \tarious. hypotheses and final1y arrived at the conclusion that "The degree of segregation in the role.,relationship of husband and w; fe var; es directly wi th the connectedness of the fami ly r s network Il (Bott. 1971: 60). She defined the social network as the people with whom a conjugal pair have soc;'l 'relationships. Th~se networks were found to vary in t~ei;degree\fconnectedness, that is, in thedegfeeto,which ,_ the peop le knownl' b (^) r a fam; 1y know and meet (^) )wi th one another ; ndependently of the fami 1y.) Sorne fami lies were' found to have a c lbse-kn^ <1- i t network. In other words, a large n.umber of the members of the network knew,

- and assoc i ated with, J^ each'^ other.^ Others^ wer,e^ found^ to^ have^ l oose-^ knit Iletworks where few relationships existeà between the members of the

,^ l < ,f ~ , (^11)

,, ,

1

pz $

t

."

network (Bott," 1971:59). The degree ofconnectednesswas found to affect the form of the role-rêlationship of a conjugal pair, such^ • tha t those wi th close kni t networks tended ta ha ve segrega ted conj uga l ~ role-relationships and those with loase-knit networks, joint conjugal ro 1es. She dealt with a number of variables affecting the form of the social network -- geographic mobility, soclal class, social mobility, educational background. for example and showed how these relate to the connectedness of the netvlOrk and, through that, to the conjugal role-relationship. Throughout her work it is" the network connectedness which affects the fonn of the role-relationshipi all other variables are factors whi ch dea 1 through the network connectedness factor ta affect the conjugal roles. In her "Reconsjdera}lOns" however she is willing to admit that these other factors can themselves alter the fonn of the conjugal role-relationshi~. There have been a number of attempts to replicate Bott's res~lts. As l shan show in the following pages, none of these have met with any degree of success. Turner (1967) ignores the dimension of kin. Udry and Hall (1965) and Toomey (1971) define networks too circumspectly. Harrell-Bond (1969) makes an untested assumption of the form of the net'lork. Both B100d (1969) and Nelson .. (1966) deal 'Iith the emotional content of the behavi our rather than its structure. In each"'case, by so doing they fail to test Batt's hypothesis adequately. Criticism shauld therefore be 1evelled both at their procedures and at the initial hypothesis. Udry and Hall (1965:392) studied the partial network of 43

-

. l

(

she naturally found no correlation between network density and conjugal rn1e-relationships. Shé concluded that the relevant factor ;5 the cultural background and the êxperience derived from the parentaJo family. Those families of Irish backgro~nd had a higher degree of marital role ~egregation than the English families. However, by assuming that all the networks were loose-knit without validating this exper~menta11y,, her results become meaningless as a replication of Bott. Robert Blood (1969:171) studied 731 married women in Detroit. He measured the jointness of the relationship by means of the wife's satisfaction wfth the husband's role performance. He concludes that his data support Bott; she herself says that it does not (Bott, 1971: 267), because he measures marital satisfaction and not marital roles. A conclusion that emerges from his paDer, however, is that conjugal role segregation increases as the proportion of relatives living in the neighbourhood increases. This will be discussed later in the chapter. Nelson (1966:663) ch9se ta define the network in terms of an individua1istic versus clique fo\m of contact rather than by de~sity criteria. His practica1 method of defining this (asking how many Deeas ions OJ1-~hich two or more of the network members met together) produces much the same results ~ a calculation of density. He diverges from Bott's method of analysis when def'ining conjugal role- relationships. His method of ana1ysis is largely based'!~n the emotional relationship between husband and wife, as the wife p~rceives it. ACQording1y, the dichotomy produeed is between socio-emotional and .

-

,^ ~,

~ <

~ i

z.

1.!

.. L

and instrumental relationships, rather than between segregated and joint conjugal roles. Although not totally equatable with Bottls criteria (which are based on activities, not emotions), there is nonetheless a certain 1 relationship tJetween the' two sets of terms. An instrumental marital relationship is defined as one in which there is little emotional involvement;' the husband fulfils his obligations to the family, but does not share an ,^ emotional involvèment.^ i^ with .his wife. This corresponds in sorne way " with Bottls segregated conjugal role-relationship. A socio-emotional relationship, on the other hand, involves a greqter involvement between husband and wife and is similar to the joint conjugal role-relationship of Bott. Nelson found that 66% of wives without clique contact thought of their husbands in socio-emotional terms, while onl~ 41% of t~ose with clique. contact did 50. He also mentions, 'in passing, that he found kin contact to' be important for the subjects. The most comprehensive study of Bottls hypothesîs was carried out by Turner (1967:121) in a rural community in England. His data ~ere initially 'gat~red for another purpose, but it was suggested that he use it 'for a replication of Bott. 1 As he was one of the few who used\ a similar method to her, i.e. an intensive stud,Y of fami1ies, Bott considers his study the most relevant attempt at replication (Bott, 197.1 :282). Nevertheless, he did not obtain the same results. Thére were those with joint conjugal role-relationship who had^ • both. dense networks and loose-knit networks. He attempted to relate the conjugal role-relationship to a number of variables: occupation,

P $

·r

<:

--

There are a number of other studies which dea1 with Bottls hypothesis, Aldous and Straus (1966:576), J. Platt (1969:287), R.

Frankenberg (1966)~ but. none- adds substantially to, the criticisms

and^ ,^ observations·.' of the above mentioned authors. Although Bottls. results were replicated by Nelson (1966) a'nd R. B106d (1969), by and l~rge her conclusions have not bee~ confirmed. This i? due to a number of factors. Nowhere does she clarify what she means by terms such as close-knit or loose-knit network, and joint and segregated conjugal roles. For example, it is not known how many members of a network need know and associate,with each other for Uott ta have classified it as a close-knit network. Udry and

Hallis (1965) three-point scale of network desnity was an att~mp~ 4

ta remedy this, but it can be argued that such a simple scale is unrepresentative of the network as a whole. Aside ·from the imprecise definition of terms, there ';s another methodological proble~ with Battis work. She,uses different units

"

of analysis at differènt times. and as a result there is sorne confusion in her conclusions. " ... in sorne parts of her work Bott treàts, the spouses separately, while in other parts she treats the married couple às a unit." (Barnes, (^). 1972:21) ... That is, when dealing' wÙh !::ouplés with segregated,~ conjugal roles she traces (^) ,out the Detworks of the individual partners, and, when dealing with couples with joint conjugal. rules, she traces out th€ network 'of^1 the couple. Her conclusions however are based upon the network of the couple -- not on the network

. of the tndividuals. Por example, she c1aims,the Newbolts have a close~ knit network on the basis of the finding ~hat~Mrs. Newbolt has·a close-

'\

r

'.

.'~^ ,^ " , .~'~) ,i

y J

-, ,0) t

knit network. A genera1ization is thus made abclUt the couple on the bâ~is of the network of one s~ause., From this conclusion abeut'the network d~l1'sity of the couple, Bott th^1 en maké'-S generalizations about the re-lat-ionship betvleen the individuals within that unit. Again. ta quote B'~ rnes : , But in the condensâtion, the relation between husband and network, wife disappears are the fTlarried from co'upJgsthe network. .... HenéeThe unitsthe Gorre1in thj ation sheand èe-rtain seeks properties to'establish of a relationis between that zone does density n'ot form' part of the1'1'e"tl'lork she is examining. (41972:22). 1 .'

Il '

In (^) ,,;o'ther words ,ô(shè/" is di scusstng (^) , the network of (^) -!one unit• -- the couple --

• ' and maHn'g '. ,: €}xt'~à~~l~tions; , J :~"" trom that toL,the behaviour of the ;nd;v;duals if W;thir. that unit.'

there have ~ beeh many a,ttempts. ta, dea 1 with, th i s PfOp lem. Turner,

"'Ih1le l"ecognisiJ:19 this diffi.culty, opts instead to deal with.. bath indrvidual and household networks and(to ignore the log;ca', problems that this p.resents. Others (Udry and Ha,l1, if •.Pla~~se only the networks of the indiviudals, and do 'not use the ma~~~:1 couple as a unit'of analYS;'S at al1. This solution ha~ problems. In 50 doing ~he res~a~c~er$ ignore the-^ , whole ~sis of a joint role-relationship,

-^ ~hich 1s/^ that the^ ,couple ~ct" as^ a^ unit^ vis-à-vis^ their^ friends and 1 relatives and not as indiviudals. A third' potentiai solution is (^) . offer€d by Nelson. , \ He deals with the netwQrk in terms of the content of the relationship. - He says: , 1 9 1 When social contacts o~ts;de the family become segmental,' indiviquals tend to be more reliant on secondarY.rather .emotional sustenance^ than^ primary^ group^ contacts. from thêir^ They spouse.^ ...^ (as a(1966:670).^ result)^ expect,

1^ J

"

..

t

.,

~^ " r" l, (:

within the kin universe. The second hypothesis,advanced to explain Bottls data is that by Turner (1967:121; c.f. a1so, Harris, 1969, Fa11ding, 19?1). This argues that it is the existence of mono-sex'networks which produce segregated conjugal role-relationships. T~rner recognised that,both spouses shared their kin. But if kin were èxcluded from the network, those with a close-knit.male network for the husband and a close- knit female neblOrk for the wife had segregated conjugar roles. That is,/^ where there is segregation outside the^1 home, there is also seg- regation wit~i~ it. Harris says "since marital ro)es are part of sex roles there is 1ikely to be a fit between mono-sex group membership and'role segreg~ion (1969:174).

MY PROBLEM My intention therefore is to examine: , ~

1. Bottls contention,that, it is the density of the network which affects the form of the conjugal role-relations;
2. G1uckman ls postulate that it is affected by a kin-based core to the social .'network, and
3. Turnerls argument^ , that the existence of mono-sex networks is the critita1 fa'ctor in the form 'of conjugal roles. 1 will examine each of these hypotheses 'against a se~ of data on the family derived from, a rev;ew of the literature. Most of the , data^ i5..^ from^ English sources,^ but^ sorne^ American^ 'studies will^ be^ used as well. ,These are largely community studies which had the family as a primary focus, but not'exclusively. rhey include.a"study of a

"

l ,

1 ",^ j' i ,^ .,a^ é'

, ," f '

, 1

...

coa1 mining town, Ashton, (Denis, et~, 1969), studies',of working class ~reas in: OXTord (Mogey, 1956), Swansea (Rosser an~ Harris, 1965),^ , Dublin (Humphreys, 1966) as wel1 as ~the famous study of Bethnal Green (Young and Hillmott, 1966). The American Jiterature is not as comprehensive, but includes studi.es of working-class wives (Komarovsky, - 1964; Rainwater, gl.li', 1959) aAd a ~ase study of an Ita1ian community. ' in Boston (Gans. 1962). For the pu\poses of this examination-^ one should rea1ise three( things. ~st. there is some confusion over terminology, and this must be qfeared up. Second, the data on families with segregated

.. rol es may ryof be represe.ntat ive of a 11 famil i es or even of a 11 fami li es with segregated roles. There may be circumstances extraneous to the .. , three network hypotheses '1hich can affect the family form, and which 1 make any such hypothesis applicable onlf te: a 'certain set of data. ~ This in turn leads to the~hird point, that the three hypotheses

\ '

d~scribed above are noticeable fo)'l'l. f thei'r.- r~arefaction. These three ignore many of the conditions beyond~the^. immediate network of the

family. The above three stat-ements,wiil p1rati'\le th~ basis of the

. final discussion . in this chapter. ., Since Bott does not define her terms precisely; and since most Di the al,lthors. with which 1 will be .deal ing were not predomi'nantly concerned./with the exploration of my hypotheses, compariion~of t'he literature is a problem. In arder to'overcome this. it is necessary/ fa~ the terms ?eg;egated conjugal role-rel~tionship. dense network: mono-sex network and kin-b~sed core t6 be defined in such à way as ta encompass, the various approaches to the data fo~nd in the literature.

JI