I am writing with regard to the recent publication in your journal concerning the acquisition, maintenance, and loss of a type of speech called a twang. Terris et al. make only cursory mention of – and thus fail to do justice to – a hypothesis that speaking with a twang might be associated with a retrovirus or another pathogen. Our lab has been pursuing this question for over 20 years and I would like to clarify the current status of the debate.
Our search for a pathogen involved in language perception and speech began with a series of observations on the phenotype: in many ways, the spread of the phenotype resembles an epidemic that is tied to particular regions. For example, Valley Fever, or coccidiodomycosis, is caused by a fungus found in dry areas of the Southwestern United States. The fungus forms spores that are spread by winds, particularly when the soil has been disturbed by storms, construction, agriculture, four-wheel drive offroading, motorbiking, or other sports activities. Inhaling the spores leads to an infection in some people.
It is estimated that about a two-thirds of the population of some regions of the Southwest will test positive for the fungus Coccidioides spp. at some point in their lives. Only a fraction develop flu-like symptoms. In severe cases, nodules form on the lungs. Their onset and their severity vary from person to person, likely for genetic reasons, which also play a role in whether the pathogen affects organs beyond the lungs. A weakened immune system greatly increases susceptibility. Symptoms may disappear and reappear over the course of a lifetime.
In many ways the spread of the twang resembles such diseases, which are caused by a pathogen restricted to a particular geophysical niche. There are “hotspots”, particularly in the Midwest, where penetrance reaches nearly 100 percent, surrounded by zones of variable penetrance. Geographical barriers may play a role in limiting its spread. The Rocky Mountains, for example, divide an eastern region of pronounced twang from western areas where it is hardly found at all. There is some evidence that following the Dust Bowl, which saw massive migrations from Oklahoma to California, the pathogen was transported to the western coast, where it was responsible for the rise of “Valley Girl” speech. It has been estimated that in their clothing and shoes, immigrants brought approximately two tons of Oklahoma dust to California. The pathogen may have come along for the ride.
Infants seem particularly susceptible; virtually every child born in a hotspot will acquire the twang, independent of his or her genetic background. Some studies indicate that the degree of penetrance is associated with socioeconomic factors. This, too, is common for pathogens associated with dirt or a lack of sanitary infrastructure. An intriguing observation comes from recent epidemiological work that links the severity of a family’s twang to the number of open beer bottles and pizza boxes lying around the house. Another correlation is the number of rusty cars parked behind the house. In each case, the higher the number, the more severe the twang.
Those exposed during early childhood typically suffer from the twang to some degree their entire lives. Interestingly, those who leave a hotspot for many years – usually decades – may lose many of its features. However, if a person returns home, for example during Thanksgiving, he or she experiences a dramatic but temporary increase in twang speech patterns. This likewise reflects the behavior of some pathogens: removed from their ideal environment, they reproduce only slowly or enter a phase of latency. Contrarily, someone who moves to a hotspot later in life may at some point begin to show symptoms, but only after prolonged exposure.
The hypothetical pathogen does not seem to be transmitted from person to person. Children raised by twang-positive parents in a twang-negative environment do not typically show symptoms. Weaker phenotypes that are occasionally observed might be explained by transmission through contact with fomites such as dust-ridden clothing, furniture, or beer bottles that have accompanied the family without being properly cleaned before a move.
The findings of Terris et al. are intriguing but do not in any way contradict the pathogen hypothesis. A range of infectious agents are known to affect CpG methylation patterns and the expression of genes. Tumors in particular regions of the brain that affect speech patterns may cause symptoms by disturbing neural networks, but they may also be accompanied by changes in the epigenetic regulation of genes.
Validating the twang-pathogen hypothesis will require studies of the metabiome of those affected compared to controls. We have recently carried out such studies using a cohort similar to the patients and controls described in the paper by Terris et al. Our preliminary work, which is currently being revised for publication, has identified three potential candidates: the strongest correlation involves a retrovirus which bears some similarity to the feline leukemia virus, and there is a somewhat weaker association to two species of fungi whose spatial distribution closely matches that of the twang. At the moment we cannot rule out combinatorial effects caused by multiple pathogens, whose lifecycles depend on a delicate balance between body homeostasis and external factors in the environment.