Mother-Infant Attachment Amongst the Dogon People of Mali: McMahon et al (2001)

• Constantly keeping infants physically close, breastfeeding on demand and responding immediately to distress are common practices within the Dogon culture but very different from our own.
• This study gives valuable insight into a culture where infants are raised by natural parenting practices and provides a contrast to the more usual studies performed on Western cultural samples.

AIMS:

• To assess whether infant attachment types are different in a culture that raises infants using natural parenting methods.
• To assess whether attachment security was related to the quality of mother-infant communications.
• To assess whether mothers of secure infants respond more sensitvely to their infants than mothers of insecure infants.

• 42 mother-infant pairs from rural Dogon villages were used as participants.
• Infants ranged in age from 10 - 12.5 months at first assessment.
• The ** testing method was used to assess attachment styles.
• Results were compared to those from 4 North American samples, with a total of 306 mother-infant pairs tested.

RESULTS:

Type A attachments:  Dogon 0% / North American 23%

Type B attachments:  Dogon 67% / North American 55%

Type C attachments:  Dogon 8% / North American 8%

Type D attachments:  Dogon 25% / North American 15%

• Many Dogon children had their grandmother as principle carer during the day but attachment classifications were unaffected by the type of primary caregiver because mothers remained closely involved with children through regular breastfeeding and co-sleeping during the night.

• Positive correlations were found between maternal sensitivity and infant security ratings and the quality of mother-infant communications and infant security ratings.

CONCLUSIONS:

• Children raised by natural child-rearing pracices have higher levels of secure attachment and no Type A attachments.  Explained by incompatibiity of Dogon child-rearing practices with Western cultural child-rearing practices associated with Type A styles. In Dogon culture there is no maternal rejection of attachments bids, intrusion or lack of physical contact.
• Naturally parented children have greater attachment security and an absence of Type A attachment, when compared to children of Western cultures.

EVALUATION:

• The high proportion of secure attachments and no avoidant attachments found among the Dogon people is remarkable considering the mortality reate of around 45% of infants before the age of 5 and the high rates of social and economic stressors e.g. high proportion of mothers with HIV and high levels of poverty.

• Findings are backed up by 2 similar studies.  Tomlinson et al. (2005) using a South African sample and Zevalkink et al. (1999) with an Indonesian sample.  Both found high levels of secure attachment and low levels of Type A.
• Comparable studies are needed of infants naturally parented in Western culture.  Rates of secure attachment should be even higher here, since factors such as poverty, social stress and child mortality are less.  One problem with such studies is the possibility that maternally sensitive mothers are more likely to choose natural parenting procedures, which could bias findings.
• The ** procedure contains elements unfamiliar to Dogon infants, like being left with strangers, thereby creating a risk that infants are being wrongly classified as insecurely attached.

Re. Q about the seemingly high percentage of Type D attachments.

• The etic may be one thing, however you can't ignore cultural differences in raising children either.
• For example, in the USA, African Americans are more likely to have several carers in comparison to their counterparts (Rogoff, 2003).

• Japanese children tend to live in the beds of their parents, if not the same room. German parents fostered autonomy.  So this could suggest there would be variances when using the strange situation with different cultures.
• Secondly, the sample sizes are not representative of their populations.  For example, the number of pairs for Britain (little island) is three times that of China (25). Only the USA tends to have a reasonably large sample.
• The mean of the meta analysis (pg 124) shows the secure attachment was pretty consistent across cultures, including that of the Dogon's (meta 65%/Dogon's 67%). So, if it were the etic, shouldn't we see a wider disparity? So the suggestion that the etic plays a strong role may be irrelevant, as the method was pretty consistent with these groups. So if it's not necessarily the method, or the etic, maybe it's the 'something else'. In the meta analysis we allude to socioeconomic, but with the Dogon it's other factors as well (mortality, disease). So, this suggests that most attachments will be similar in outcome (secure), but external factors will impact on the quality of the other attachments. 
• So, secure attachment looks (ready for it) secure, but in our differences there are variances. You will find where there are differences in attachment, there will also be differences in future relationships as well.