Yale anthropologist: Muddled terms hinder study of monogamy in mammals
Fuzzy terminology, faulty methods, and funky data have plagued recent scholarship on the evolution of monogamy among mammals, according to a pair of studies co-authored by Yale anthropologist Eduardo Fernandez-Duque.
Fernandez-Duque and his co-authors, Maren Huck of the University of Derby in the United Kingdom and Anthony Di Fiore and Sarie Van Belle of the University of Texas-Austin, illustrate their concerns with numerous examples, including two large-scale 2013 studies that attempted to provide a comparative analysis of the evolution of monogamy: one broadly focused on all mammals and the other concerned with primates.
“These kinds of large comparative studies have flourished as it has become much easier to access data from various sources,” said Fernandez-Duque, professor of anthropology at Yale. “Studies that address thousands or hundreds of species seem to attract more attention from top journals and the media than more focused research does, but unless they use reliable, high-quality data, the results will be questionable.”
The first critique is published in the journal Frontiers in Ecology and Evolution. The second is forthcoming in the 2020 issue of the Yearbook of Physical Anthropology.
The co-authors have multiple decades of experience researching the mating and social systems of four monkey species — owl, titi, tamarin, and saki — in parts of Ecuador, Peru, and Argentina. Owl monkeys and titis are consistently found living in pairs, whereas sakis and tamarins have more flexible social systems.
They show that the 2013 analyses use imprecise, inconsistent, and interchanging definitions of important terms like “pair living” — which describes a social organization — and “sexual monogamy” — which describes a mating system — sowing confusion and creating “apples and oranges” comparisons.
The terminology used needs to be clear and explicit,” said Fernandez-Duque. “Different researchers are using different terms and criteria, meaning that comparisons between studies are essentially meaningless.”
Comparative analyses of pair living and sexual monogamy are relying on inconsistent and low-quality data, according to Fernandez-Duque and his co-authors.
“We’re of the slow-and-steady school that believes comparative analysis requires examining primary sources based on the original data acquired in the field,” he said. “We see studies using data from encyclopedias and other secondary and tertiary sources. Little attention is paid to the quality of the data being used in those studies.”
The critiques suggest that the technology enabling easy access to sources and data is harming the quality of research. A search of the scholarly literature can produce dozens of citations on a given species, but those references frequently don’t include primary sources, Fernandez-Duque said.
“You cannot determine whether a species has been adequately studied by counting citations,” he said. “What if the studies cited are of poor quality or based on animals in captivity rather than in the wild?”
To illustrate this point, Fernandez-Duque described how the large-scale comparative analyses from 2013 examined the “infant-care hypothesis,” which posits that monogamy evolved among certain species because parents mutually benefit when the father provides care to their offspring. Any evaluation of the hypothesis must carefully account for whether paternal infant care is present, and how, within each species, he said.
Fernandez-Duque and his colleagues determined that two-thirds of the references cited for paternal care in one of the studies were inadequate, and all but two of the references in the other paper could not be reconciled with the primary source.
Fernandez-Duque and his co-authors conclude their critique with a plea to return to “the bedrock” of scientific inquiry: the collection of high-quality natural history data on a wide range of animal species. They call for distinguishing between conclusions based directly on high-quality primary data and those founded upon statistical inferences and creative thinking.
“While creative thinking can open up new avenues of research, the quality of science ought to be defined by solid data and rigorous methods,” he said.