Sometimes, people suggest that languages are related just because of a small number of lexical similarities between them. However, it could be extremely likely that these words appear similar by sheer coincidence. Perhaps, one of the most well-known examples quoted is the rather striking similarity between English and Mbabaram, for the word “dog”.
However, English and the extinct Mbabaram are far from remotely related — English is a Germanic language, and Mbabaram a Pama-Nyungan language, branches that might have split off way back into the past from other ancestral languages. How “dog” was used to mean “dog” in both languages is a sheer coincidence. In fact, it is possible that this word originated from independent etymologies. True cognates for Mbabaram like Yidiny gudaga, Dyirbal guda, Djabugay gurraa, and Guugu Yimidhirr gudaa show shared word origins, all of them most likely originating from Proto-Pama-Nyungan *gudaga. However, in English, the word “dog” has an etymology that is shrouded in mystery. While its Old English form was docga, further poking into how this word arose often yields in uncertainty.
This example is what is known as false cognates — a pair of words that appear to be cognates due to similar sounds and meaning, but in fact, have different etymologies. This should not be mixed up with false friends, where two words in different languages or dialects look or sound similar, but have different meanings.
False cognates extend far beyond the comparison between English and Mbabaram. In fact, even in more closely related languages like Spanish and English, some false cognates could be seen. Another example is the differing etymologies of the words much and mucho in English and Spanish respectively. While the former derives from Old English myċel, ultimately being reconstructed as *meǵa– in Proto-Indo-European to mean “big, great”, the latter derives from Latin multus, reconstructed as *ml̥tos in Proto-Indo-European to mean “crumbled”.
This also brings us to the seemingly close-sounding words for “thank you” in Japanese and Portuguese. In Japanese, this translates (in informal speech) to ありがとう arigatō, while in Portuguese, this translates to obrigado. The former derives from the word 有難い “arigatai” to mean “thankful” or “appreciated”, while the latter derives from the Latin word obligātus to literally mean “obliged”.
Even within English, one could find false cognates. This derives from words having similar sounds and meanings, but these words have their own different origins. Typically, these words could find one of their origins in Old English, while the other could be derived from Latin origins. One example are the words isle and island. While both words refer to the same object depicting a piece of land surrounded entirely by a water body, the word isle derives from the Latin word insula, while island derives from the Old English word īġland.
How about other special cases of false cognates? The basic kinship terms for “mother” and “father” often share unusually similar sounds, such as “mom” and “mama” for “mother”, and “papa” for “father”. Examples include Vietnamese, with mẹ meaning “mother” and bố meaning “father”. Linguists have proposed that these words could derive as a result of this process known as early language acquisition, basically the process where a young child or infant picks up their parents’ language or languages. Babble words or sounds have been widely put forward as an explanation for this coincidences in sounds, as they are observed to be the first words or sounds articulated by babbling babies. Parents would then associate the first sound babies make with themselves, then incorporate them into their baby-talk lexicon.
While the etymologies here are heavily simplified, it still stands to show how words, no matter how similar they are in meaning or sound, have their own different stories behind them, but have since converged. The wider implication here is to warn against deciding that languages are related based on this small number of lexical similarities.