Words in English public website
Ling/Engl 215 course information
Rice University
Prof. S. Kemmer

Parsing


To parse a word means to analyze it into component morphemes. Recall that morphemes are the smallest units in a language that link a form with a meaning or function.

Parsing is generally done on complex words that came from Latin and Greek. (We call such words "Latinate" vocabulary or "Classical" vocabulary.) Such words typically show the clearest word structure, in part because Latin and Greek had many affixes for inflection and for dervivation, and (unlike in Germanic) their word structure REQUIRED putting together roots with affixes. Further, many Classical words were coined long after the classical period, so the word structure is more transparent than words from English or French that have been in the language so long that their morphological structure has become murky. With many native and nativized words, what were once separate morphemes have over a long period of time fused together.

For example, the native word stirrup comes from stig 'climb' + rap 'rope'. The word meant in Old English 'loop of rope for placing the foot to climb on a horse'. This word was a compound in Old English, with two separate morphemes, but now it is a single, unanalyzable morpheme with the modern meaning 'device for holding the foot when mounting and riding a horse.' The whole word now has one morpheme instead of two, and it no longer refers specifically to rope at all.

Parsing Practice

The following example words are for parsing practice. For each morpheme in a word:

1. specify the form of the morpheme (the major allomorphs, separated by slashes)

2. below it write the morpheme's meaning or function. (There may be some parts of the word that are "linking forms" without any meaning.)

3. To complete the parse, we state the actual meaning of the whole word in Modern English. Note that this meaning may be somewhat indirectly related to the component morphemes.

Examples:


egregious
e/ex  +  greg           +   i         +     ous 
out     social group      (linker)          ADJ
'unpleasantly standing out; blatantly bad'  

azoic
a/an   +        zo      +       ic
not           animal            ADJ
'without (animal) life; preceding the appearance of life on earth'

creation
cre         +        at(e)     + ion
'create'             V           N
'the final product in the process of making something new'

More etymologically exact parse:
creation
cre         +        a            +  t(e)     +        ion
'create'             stem vowel      Past. part.       N
'the final product in the process of making something new'

The 'e' in parentheses is only there for spelling reasons--it has no etymological connection with the word for 'create' in Latin. It is only a prompt to remind us that the morpheme /ate/ is pronounced with a front mid vowel.

Sample words for parsing

One set of sample words comprises the phonetics terminology for our class. For these words see Sound terminology.

apteryx              hippopotamus        megalith

perihelion           bilabial            eliminate

transliterate        seminal             iatrogenic

anhydrous            biennial            apnea           

endoscopy            supercilious        aphelion            

inculpate            exophthalmic        laryngoscope

anemia               osculate            subcutaneous

luminary             amygdala            polysemy

pandemic             androgynous         agenda

memorandum           exculpate           hippocampus

More sample words for parsing

confluence      megalith     incarnation

cryptogenic     geminate     phyllophagous

nyctitropism    phototropic  phytogenic

aphasia         perigee      oenophile

formicivorous   apterous     aliform

arachnophobia   apiculture   oology

galactic        errant       errand

Parsing vs. Etymology

Parsing is related to finding the etymology of a word, but it is a little different because the focus is on word structure, rather than word history. This has various consequences.

Word structure (for our purposes) includes primarily roots and affixes. So, many of original bits of the source word, such as inflectional morphology in the original language, are not relevant to a parsing.

For example, for hippopotamus, you mind find in a dictionary etymology that the word comes from Greek hippos 'horse' followed by Greek potamos 'river'. The dictionary etymology might also indicate that the -us ending comes from Latin (Latin and Greek were fairly closely related languages, and the Greek noun inflectional ending -os is historically/etymologically the same as Latin -us.)

In a parse, we leave out the information about what language the word parts come from: it is not relevant for this purpose. Even more important, we also strip the source elements down to their roots, removing inflectional endings from the original language that the dictionary etymology included, if they do not survive in the borrowed word.

The resulting parse:

 
hippopotamus
hipp     +     o            +   potam         +     us
'horse'       linker            'river'              'noun inflection'
'large thick-skinned herbivorous mammal living in and around tropical waters
of Africa'.

For the definition, you have to get close enough to the modern meaning for someone to understand the thing defined as something distinct from similar things, but you do not need a very technically precise definition. For our purposes 'large African mammal living around rivers and swamps' would be good enough.

Important: Definitions should preserve the part of speech of the word defined. So you would not define somnambulant as 'to sleep-walk', but rather 'sleep-walking'. It is an adjective, not a verb, so the definition must be appropriate for an adjective.

As stated above, parsing is generally done on complex words that came from the classical languages. The aim in parsing is to find out the structure of the word, isolating the meaningful elements that recur not only in this word but in other words, so that we can learn more of those elements and learn more words that use them.

Etymology, on the other hand, is more like the story of a word from the earliest point we can trace, to its modern meaning. Etymology can be done on any word, because all words have SOME history. Even a novel creation like googol 'mathematical term for 10 to the 100th power' has an etymology: "Novel creation of amusing-sounding word by young son of the mathematician who defined it". But it wouldn't make too much sense to try to parse googol, because it is a simplex word, i.e. it has only one morpheme in it.

In the hippopotamus example, the parse is different from the etymology, not only because a parse does not include the source language of loanwords as an etymology does, but also because some dictionary etymologies break the word down into whole source words instead of roots, e.g. an etymology might state: "from L. hippopotamus, from Gr. hippos 'horse' + potamos 'river' " . (Dictionary etymologies are heavily abbreviated and you have to figure out the abbreviations for the dictionary you use.) The -os part of both of the components of the compound was just a Greek inflectional ending signalling a certain class of masculine noun with nominative case. It's not in the parse because it doesn't show up in the word today. The -us ending of hippopotamus, on the other hand, DOES show up in the modern word so we must take account of it. In fact it is the Latin version of the same Greek inflectional ending seen in hipp-os. It is enough to just gloss it as 'noun inflection'. Later (Ch. 9) we will learn some of the inflectional categories of Latin and Greek which have ended up in our English words.

To find the elements relevant to parsing, look in our textbook in Appendix 1, starting on page 221. These elements are the pure roots and affixes, without additional morphology, such as inflectional morphemes that allowed them to be used in whole words in Latin and Greek. That is what we want to use in parsing: roots and affixes.


© 2003-present Suzanne Kemmer
Last modified 23 Aug 13