The Comparative Method IV

Since our whole analysis rests on regular phonological correspondences, let's take a closer look at what RPCs are and how to recognize them.
You can't find a valid RPC by only examining a handful of word sets; however, if you don't have a particular RPC to begin with, how do you look for more examples of one? The resolution of this catch-22 is a sort of bootstrapping method - at first, you do have to start with a handful of word sets; once you start recognizing likely patterns, you need to find more sets that will confirm or disconfirm the likely pattern.
In the European set we've been examining, I divided the six languages into two subgroups based on a few correspondences. The English-Dutch-German subgroup is an actual family called Germanic - it includes not just these three languages, but also Yiddish, Afrikaans, Frisian, Scots, and the Scandinavian languages Danish, Swedish, Norwegian, Faroese, and Icelandic. The French-Italian-Spanish subgroup is also a family, one which many of you may be familiar with - the Romance languages. This group includes the three just mentioned plus Portuguese, Catalan, Provençal, Walloon, Romansh, Sardinian, Sicilian, Romanian, and many others, all of which are descendants of a language known as Vulgar Latin, which was the colloquial language of the Roman Empire, alongside the Classical Latin language used in Roman government and later in the Catholic church. These two varieties descended from an archaic Latin language that was spoken in central Italy, along with sister languages that included Faliscan, Oscan, and Umbrian. This family is called the Italic languages, and includes the Romance languages as a subfamily. Both the Germanic family and the Italic family in turn are descended from an ancient language that linguists have called Proto-Indo-European, which was spoken before 3000 BC somewhere in the region between Asia Minor and the Ural mountains (I'll talk more about this and other language families in a future post).
Anyways, was there enough data to justify me placing the six languages in our set into these two groups? So far, not really. But there were some promising RPCs that do provide a place to start. Let's look at the table again:

English	Dutch	German	French	Italian	Spanish	gloss
aɪ	ɪk	ɪç	ʒə	ˈio	jo	‘I’
mun	maːn	mɔnd	lyn	ˈluna	ˈluna	‘moon’
ˈsɪstəɹ	ˈzʏstəɾ	ˈʃvɛstəʁ	sœʁ	soˈɾɛlla	eɾˈmana	‘sister’
ˈfaɪəɹ	vyːɾ	ˈfɔʏəʁ	fø	ˈfwɔko	ˈfweɡo	‘fire’
eɪt	aχt	aχt	ɥit	ˈɔtto	ˈoʧo	‘eight’
fʊt	vut	fuːs	pje	ˈpjɛde	pje	‘foot’
dɹɪnk	ˈdɾɪnkən	ˈtʁɪnkən	bwaʁ	ˈbeɾe	beˈbeɾ	‘drink’

We see the fvf correspondence in Germanic initial consonants in two word sets, 'fire' and 'foot'. And as we saw in this table:

English	Dutch	German	French	Italian	Spanish	gloss
ˈfɑðəɹ	ˈvadəɾ	ˈfatəʁ	pɛʁ	ˈpadɾe	ˈpadɾe	‘father’
fɪʃ	vɪs	fɪʃ	pwaˈsɔ̃	ˈpeʃe	pes	‘fish’
foɹ	voːɾ	fʏʁ	puʁ	peɾ	poɾ	‘for’
fli	vlo	flo	pys	ˈpulʧe	ˈpulɡa	‘flea’

the same correspondence can be found in several more word sets, along with the equivalent ppp correspondence in Romance initial consonants for the same words. Now, how do we know that the ppp set is the one that corresponds to fvf, and not the fff set we see in the Romance words for 'fire'? We don't at first. But looking at more items, such as the ones in the second table, we will eventually find many more sets where ppp corresponds to fvf than where fff corresponds to fvf. This suggests that the fvf-fff set is not a real correspondence, but is merely a coincidence.
So assuming we've confirmed that fvf-ppp is a valid RPC set, we can begin to look for more examples. The more we find, the surer we are of the validity of the set. Here are a few for the Germanic languages:

English	Dutch	German
fi	ve	fi
fɝst	vɔɾst	fʏʁst
faʊl	ˈvoːɣəl	ˈfoːɡəl

We see the same fvf correspondence set. But this time there's a snag - the words in each set do not mean the same thing. In the first set is the English word fee, but its Dutch and German cognates, vee and Vieh, mean 'livestock, cattle'. In the second, we have English 'first', whose Dutch cognate (vorst) means 'ruler, monarch' and whose German cognate (Fürst) means 'prince'. And in the third set, we have English 'fowl' next to the Dutch and German words for 'bird', vogel and Vogel.
So, are these still cognates? Yes. Cognates are defined by their common ancestry phonologically; that is, they have evolved from their common ancestral word through a series of phonological mutations. But, human language being how it is, words also undergo semantic changes. Cognates can still be cognates even if they mean completely different things. And even so, in many cases the meanings don't have to be that different. Obviously, there's a semantic connection between 'fowl' and 'bird', the first being a specific type of the latter. Based on the fact that the Dutch and German cognates both mean 'bird' (a sort of ‘majority rules’), we can conjecture that this was the original meaning of the ancestral word, and that the English word 'fowl' at one time could refer to any type of bird. When we look at the word for 'bird' in other Germanic languages, we find the same cognate, which bolsters our conjecture that ‘bird’ was the original meaning.
Now I leave two exercises for the reader:
1 What semantic connections can you see between 'first' and 'monarch/prince'? How about between 'fee' and 'livestock/cattle'?
2 Look at the following table (the data is constructed). Are there any regular phonological correspondences that seem promising? How would you group the languages in this table based on them?

A	B	C	D	E	gloss
ˈhate	kaas	fuu	ˈkati	ˈputi	‘wind’
daˈpæɡe	niˈɾaax	ʦəˈbee 1	taˈpeke	niˈleɡa	‘mountain’
bɔˈdoja	moˈnuul	məˈnuur	moˈnula 2	moˈnaula	‘heart’
beˈzaɡa 3	piˈɾaax	fəˈɾaa	piˈɾaka 4	piˈlega	‘girl’
daˈɡoɡo	hoof 5	ʔoo	ˈhopu	taˈkuɡo	‘sea’
deˈɡeli	teˈxiil	ʦəˈɡii	teˈkili	teˈkajli 6	‘mother’

Semantic irregularities:
1 this means 'hill' in language C
2 this means 'liver' in language D
3 this means 'wife' in language A
4 this means 'sister' in language D
5 this means 'lake' in language B