A Monthly Column for Word Lovers
Destiny, Biology, and Vocabulary
A recent article in the Washington Post (warning: long read) gave a fascinating account of how DNA testing led to the discovery of two identities switched inadvertently at birth, a century ago, that resulted in two children growing up in families that were a mismatch for their actual ancestry—in this particular case, Jewish and Irish. It's not the first surprise that direct-to-consumer DNA testing has brought to light and it will certainly not be the last. From a linguistic perspective, the now burgeoning field of genetic genealogy provides an interesting case study for the ways in which we develop new terminology for new concepts, picking and choosing among the raw materials we have (that is, words) to designate things for which we didn't have or need particular names before.
It is our instinct, when new things come along, to give them names that form a helpful connection to something we already know. The two natural ways to do this are metaphor and retronym—the name for a term we create to make a distinction in language that was previously not necessary. For example, we may now use the retronym "dairy milk" to distinguish milk that comes from cows, formerly just called "milk," but now in competition with almond milk, soymilk, rice milk, flax milk, hemp milk, and the like. None of these last four is really milk, but they settled naturally into a metaphoric extension of "milk," drawing on some of its qualities: opaque, whitish liquid that's always in your fridge for drinking, pouring on cereal, adding to coffee, etc.
A popular retronym in genetic genealogy (henceforth: GG) is DNA relative. How's that different from an ordinary relative? Not different, really, but with DNA testing it became desirable to have a term to designate a blood relation that you learn about only from the fact of your sharing a certain amount of DNA. Of course, the people you call your mother, father, siblings, and so forth are probably also DNA relatives—but you don't have molecular evidence of this until you have tested yours and their DNA. Here's a screenshot of my top DNA relatives from one of the popular testing companies, 23andMe:
On this website and others like it, lists like this are popularly called "matches," and that's a perfectly straightforward new use of an old word. Someone who shares DNA with you is a match because one or more segments of DNA in your genome exactly matches one in theirs. The upshot is that you are necessarily biologically related via a common ancestor—whom you may or may not discover, depending on the quantity and accuracy of documented family history on your families.
The first three matches on the list above are other people in my family who have also been tested, and the relationships noted in the second column—those predicted on the basis of the amount of DNA we share—are in fact the relationships I have with these folks according to the paper trail. The fifth match, obviously, chooses to remain anonymous. The fourth match, "WP,"; is a surprise relative, someone I don't know, and the sort of match that typically comes up when you do one of these DNA tests.
What if you had your DNA tested and you discovered a surprise DNA relative much closer than a second cousin? For example, a parent, uncle, niece, or half-sibling you'd never heard of? It happens with moderate frequency in DNA testing, and it is usually evidence of an NPE, a term that is now standard in GG to telegraph non-paternity event. An NPE is "when someone who is presumed to be an individual's father is not in fact the biological father." This, along with known adoption, is nearly always the explanation for adventitious DNA relatives within three degrees of you. NPE is not the term I would have chosen as the standard for this phenomenon; it seems to imply that paternity didn't happen, but clearly, where there is a birth, there is paternity! "Misattributed paternity," a competing term, would be a better choice to describe this phenomenon.
The task of tracking down your connection to a surprise DNA relative can be a compelling one, especially for adoptees or others with nagging doubts about their birth parents. That's another retronym and one that was around before GG, but it has gained a lot of currency from the ever-growing number of people who are delving into their ancestries with the assistance of DNA testing. Here's a Google Ngram showing the dramatic increases in the use of birth mother/father/parent, starting in the late 20th century.
As the quest to determine birth parents has gained momentum, new tools have been developed for the task, and along with those new tools, new terminology also comes into use. The greatest concentration of these tools can be found on a website called GEDMatch.com, where people who have had their DNA tested by a number of different companies can upload their results for comparison and thus greatly increase their pool of possible matches.
One of the interesting tools on GEDMatch is called Lazarus. You remember that guy: in the New Testament account, Jesus raised him from the dead, four days postmortem. The genetic tool called Lazarus borrows the metaphor of resurrecting the dead. It enables you to make a composite profile of a dead parent's DNA, provided that you have results of your own and your other parent's DNA, and some DNA matches from your dead parent's side of your family—something that you are very likely to find exists on GEDMatch.
Another tool on GEDMatch is called triangulation. For genetic triangulation you don't need a compass or surveyor's tools. The term, borrowed from surveying and mathematics, is remarkably apt for genetic genealogy because it remains true to its root of "three." Triangulation in GG is a way of confirming definitively whether the apparent shared DNA of two people actually points to a common ancestor (the third person) for them. Just as with old-school triangulation, it is nearly impossible to define genetic triangulation succinctly; there's a good description of it here.
The point of triangulation is to find, if possible, the MRCA of the two people you're looking at. The what? MRCA is a handy initialism for "most recent common ancestor." A mark of any field of endeavor is its jargon, and the impulse to reduce frequently used, unwieldy terms to shorter versions. GG is riddled with such initialisms such as NPE and MRCA, as well as a new spin on BF: it's not boyfriend in GG circles, it's birth father.
Perhaps the handiest set of initialisms in GG is the shortening of kinship terms, which English has a way of designating precisely, though in a long-winded way. First cousin twice removed? Why not just say 1C2R? And so on with other cousins removed that you are likely to discover in the shared strands of your genome. This chart shows the amount of DNA that you are likely to share with various relatives. It takes advantage of the limited space available by abbreviating all the combinations of cousin + generational removal.
The other frequent abbreviation in the table, cM, is for centimorgan, a unit of measurement used for estimating lengths of DNA. It represents yet another method of coinage: the eponym. It's named for Thomas Hunt Morgan, an American geneticist.
Perhaps the greatest semantic challenge presented by DNA testing is the nuance that it places on words like ethnic and ethnicity. Such words are typically defined with reference to culture, religion, language, and customs. So if you are, genetically speaking, an Irish baby, and you happen to be raised in a Jewish family, what is your ethnicity?