Fortson Chapter 4
- Last chapter was about phonemes, sounds.
- When a sound or group of sounds change in a language, all
the sounds in a certain environment change. All of them.
- This chapter is about morphemes, units of language that carry
meaning.
- morph-eme, and hence morph-ology
- morphology is the study of the forms in a language that
have meaning, specifically their shape and how they are
'formed' (Greek morph means "shape").
- They can be a word, like 'word', or they can be a part of a
word, like '-s' in 'words.'
- They change too.
- But not across the board like sounds.
- When we change over to using "sneaked" instead of "snuck" or
vice versa, that doesn't mean that "creaked" will ever change
into or ever was "cruck."
- Changes to morphems are called morphological changes.
- Morphological change happens haphazardly, because
- a word is somehow similar to another word and so people
switch it to that word's pattern (called "analogical"
change: it is done 'by analogy' with another pattern or
word).
- when this sort of change happens in a more
across-the-board way, it is called 'paradigm leveling'
- but it often skips over the most common words, which
preserve their old forms, because people like their
commmon words to stay the same
- people come up with some explanation of the shape of the
word and that makes the word change (called 'folk
etymology').
- speakers all change how they do something, like refer to
'them, their, they'
- In English, we use Old Norse-derived pronouns 'they,
them, their': that is mind-blowing; why would
speakers change pronouns, which seem so very
basic?
- In terms of trying to find the morphological pre-history of a
language, one looks for the exceptions, the rare forms, because
they are likely to contain fossils from previous times,
especially if they are common words (it is easier to preserve
old irregular forms if you use them a lot than it is if they are
rarely used).
- We speak of "roots" of words. What are they?
- In this course, they are the most basic elements of a group
of words that are usually common to the whole group and make
it that case that we can reasonably call it one word.
- For example, 'run, ran, running,' and even 'runned'
- r_n is common to all those words.
- But r_n is not a word itself.
- We can add V to indicate that there is a vowel there: rVn
- So we might speak of rVn as the root of all those words
above.
- So what did the PIE root look like?
- Look at Fortson 4.2
- PIE roots typically had Consonant-vowel-consonant: CeC
- but they could add in resonants on either side of the
consonant CReC/CeRC/CReRC
- and they could have a sibilant (/s/) before the first
consonant. sCeC or with resonants added near the vowel.
- compare *steg > stegosaur (via
Greek), toga (via Latin), integument
(via Latin).
- And a few started with two stop consonants, a dental and a
velar: called "thorn clusters"
- called that because:
- almost all the daughter have the order velar-dental, and
the dental is /th/ as in 'thin', which is symbolized by
the letter 'thorn' in some alphabets: hence "thorn
clusters"
- chthonic, phthisic (cf.
Sanskrit क्षिति
kṣiti decay, ruin)
- only Tocharian and Anatolian have the order
dental+velar: held to be oldest, and so the others
suffered metathesis (switching of the order of things).
- Before laryngeal theory, PIE linguists thought there were cV
and Vc roots, and even VRVC roots!
- But laryngeal theory changed those into cVh and hVc roots,
which are the same basic CeC Consonant-vowel-consonant
structure as all the others, as explained above: that makes
for elegance and consistency, and we all like that, don't we?
- *h2ent- > *h2ant- > *ant-
- *h3ed- > *hrod > *od-
-
*h₃éd-e-ti (thematic root present)
- *h₃éd-ye-ti (ye-present)
- Hellenic: *óďďō
- Ancient Greek: ὄζω (ózō)
- Doric Greek: ὄσδω
(ósdō)
- *h₃e-h₃ód-e ~
*h₃e-h₃d-ḗr (perfect)
- Hellenic: Ancient Greek: ὀδώδειν
(odṓdein)
- *h₃od-méh₂
- *h₃éd-os ~ *h₃éd-es-os
(perhaps)
- Armenian: Old Armenian: հոտ (hot)
- Armenian: հոտ (hot)
- Italic: *odōs
- Old Latin: odōs
Latin: odor
- *h₃ewis > *h3owis
> *owis (cf. Latin ovis "sheep," Sanskrit अविला ávilā
“ewe”, Greek ὄϊς)
- Luwian has ḫawi- (which points to *h2e, but the
difference may be accounted for by ablauting)
-
Tocharian B: awi
(pl.) (< *h₂ewéyes)
- Armenian: հովիւ (hoviw, “shepherd”) (< *h₃ewi-peh₂
< *h₃éwis + *peh₂- (“to protect”))
- so now a lot of PIE folk write all the VC roots as HVC even if
there is no comparative evidence to justify a laryngeal: that's
called paradigm leveling: here it's done by linguists, but in
the real world, it happens all the time in languages.
- The VRVC words have changed with laryngeals too:
- Such words used to be described as having a prothetic vowel
that was unaccounted for. Laryngeals account for it.
- Greek changed laryngeals as follows: *h1 > Greek e, *h2
> Greek a, and *h3 > Greek o
- *h1reudh- > Greek eruth- "red" (cf. erythrocyte),
*h2ner- > Greek aner "human", and *h3mei´gh-
> Greek omeikh- "urinate" Armenian meg
"cloud"
- That's called the "Greek triple reflex" and is part of
what lies behind the standard idea that there were three
laryngeals.
- There are also some roots with laryngeals that have the
pattern HCeC or CeCH (some with resonants thrown in too).
These are called
seṭ and aniṭ roots in Sanskrit grammar, which
means "without i" and "with i" because i is the Sanskrit
reflex of a vocalized laryngeal, but note that in the PIE
roots of this sort, it doesn't seem that the laryngeal is
vocalized: *dheh1s- "put"
- FUNDAMENTAL VOWEL
- Usually, PIE has e as the default vowel of a root,
but sometimes there is reason to have an a, because
they ablaut differently than the rest (more on that soon).
- BITS and BOBS added here and there
- 4.10 There are some roots that sometimes have bits added for
no discernible reason: *steu-k, *steu-g, and *steu-d-
can also be *teu-k, *teu-g, and *teu-d-,
from which English stoke, Greek túkos "hammer"
and Vedic tudáti 'beats' are derived.
- ABLAUT!
- Vowels change, and they change according to a neat pattern.
- No point retyping it: see P. 79 in Fortson, where you can
find a nice illustration of the several grades of ablaut in
the root *sed- "sit"
- SO a root can have a long o, a short o, a schwa, nothing, a
short e, or a long e.
- Why? So that languages can form various things from it:
different verb forms, nouns v. verbs.
- Importantly, ablaut applies to not just roots. Suffixes too.
- And there is a different rare ablaut pattern: roots like *sal-
"salt" (and the word "salt") can ablaut to zero-grade *sl-
(as in English "silt"): and
*néh₂s > *nas-
"nose" can ablaut to *nās as in Latin nāres "nostrils"
- Word structure
- PIE words consist often of a root + a suffix + an ending
- R + S + E
- all three can ablaut!
- changing the ablaut, or the stress, of a word can produce a
different word
- Think of English surrogate v. surrogate or
minute v. minute
- This process of deriving a word from a word within the
same language is called internal derivation.
- Reduplication: mamas and papas and poopoo and peepee
- Baby talk in English, but not in PIE (or Hawaiian)
- kind of like Pig Latin: take the first consonant and
repeat it with a vowel, as in Greek tithemi or gignomai
- Inflection
- Think of English "they, them, their, theirs"
- Or "sing, sang, sings, sung"
- These changes to a word are called "inflections" (and they
also happen to have ablaut variants)
- In all PIE languages, inflection was the rule: verbs and
nouns and pronouns and adjectives all inflected, a lot!
- That is why they needed Root + Suffix + Ending (R+S+E),
whereas modern English has pretty much done away with a lot of
that and substituted word order for many of the functions of
inflections.
- A lot more on this to come.
- Sometimes inflection was done by adding a vowel after the
root, and then adding endings. Those vowels are called
"thematic" vowels and the inflections done with them are called
"thematic inflections".
- Sometimes there was no such vowel, and the inflections done
without them are called "athematic inflections"