Linguistics 300, F12, Undergraduate tutorial

Linguistics 300, F12, Assignment 9

Assignment
Spreadsheet with French loanwords (1 Nov 2012 | 12 Nov 2012 | 26 Nov 2012 | 28 Nov 2012 | 3 Dec 2012)
- We will add updates and corrections to the spreadsheet as necesssary; please keep this in mind as you plan your work for the assignment.
Notes on spreadsheet headings
Introduction to IF statements
Sample work (with calculations using IF and COUNTIF)

Your assignment for the word stress project is to explore some question (or set of related questions) concerning the change in stress assignment in our sample of French loanwords borrowed into English from about 1066 to 1620. The sample you'll be working on contains data collected in previous years. The data that you've collected this year has not been added; Akiva and I will review those words, and they will be included in next year's spreadsheet.

In framing the questions you explore, recall that the focus of the project is not a description of the French loanwords themselves, but rather of the mental processes of the speakers acquiring those words, as they are accessible to you in the attested stress patterns.

The data in the current spreadsheet have gone through several cycles of revision; however, as is usual in projects of this scope, you should expect occasional changes and additions. The link to the spreadsheet at the head of this page always points to the most current version, and you can easily identify any updates by sorting on the "Corrections" column.

The same questions can be explored by more than one student, and you are welcome to discuss the topics among yourselves, but when you write up your results, you should do so independently. Some of you will have no difficulty coming up with questions to explore, but if you are having trouble with ideas, please don't hesitate to make an appointment to discuss your project with Akiva or me.

Notes on spreadsheet headings

Corrections Sort on this column to locate updates and corrections.
Investigator Same as in Assignment 6.
Class
Headword
Usage
OED date
Effective date Here are the guidelines that I've used to map the dates in the "OED date" column into the computationally tractable form in the "Effective date" column.

circa (for instance, c1349)
A "circa" date is the best estimate that the OED editors can give regarding the date of first attestation. In reducing such a date to a number, I ignored the "circa."

ante (for instance, a1250)
"Ante" (Latin for 'before') means that the loanword is believed to have been first attested before the date given. The problem is that there is no indication of how many years before. In assigning "ante" dates to time periods, I ignored the "ante" and subtracted a fixed number of years from the remaining date. I chose 25 years (the "offset" in the last column of the header row of the spreadsheet). You are free to experiment with this parameter.

date ranges (for instance, 1305-6)
I took the arithmetic mean of the range and rounded to the next year.

implied date ranges (for instance, 135. or 14..)
I took the (rough) midpoint of the implied range. In other words, I treated 135. as 1355, and 14.. as 1450. If there was a known date earlier than the midpoint of the range, I used the known date.

question mark (for instance, ?1349, ?a1100, ?c1475)
I ignored any question marks and treated the remainder according to the preceding guidelines.

Etymon Same as in Assignment 6.
Syllables
Chaucer stress, verse-medial Stress is recorded using the same conventions as for the "OED stress" column, augmented by the various considerations laid out in Assignment 8 (especially in item 3).
Chaucer stress, in rhyme
Chaucer comments
Shakespeare stress
Shakespeare comments
OED strees Same as in Assignment 6.
Prefix 1 Information concerning a headword's morphological structure. Useful for testing various hypotheses. For instance, did English speakers categorize (at least some) prefixes as on a par with Germanic unstressable prefixes? Another hypothesis: Words containing certain suffixes (notably, -ion, -Vble, -Vnce), might be borrowed from French or from Latin, and speakers wouldn't necessarily know which language was the source language. As the numbers of loanwords from Latin went up during the Renaissance, speakers might have recategorized loanwords from French with these suffixes as loanwords from Latin (even if there is incontrovertible philological evidence that they were originally borrowed from French) and treated them in special ways.
For some suffixes, I distinguished two forms (for instance, "Vnce" and "Vnce 2"). The "2" means that the root of the headword is itself an English word. The absence of a "2" indicates that the root is a "cran" morpheme.
Prefix 2
Suffix 2
Suffix 1
Phrasal etymon? Was the etymon of the word a French phrase (rather than a simple word)? Mostly for my convenience, but feel free to use if helpful.
Initial epenthesis? Mostly for my convenience, but feel free to use if helpful.
Final syllabic C in etymon
Consistent with Quantity Stress Rule? As you will see, it turns out that many French loanwords are exceptions to both the French and the Germanic stress rule. This column encodes whether the stress on a word is predictable from a third stress rule - what I will here call the Quantity Stress Rule and what is generally known as the Latin Stress Rule. According to this rule, word stress falls on the first syllable of a disyllabic word. In words with more than two syllables, stress falls on the penultimate if it is heavy, and else on the antepenultimate.
It is usually clear whether the penultimate is heavy, but not always. In particular, it is not clear where the syllable break falls in certain consonant sequences, such as geminates, affricates, and clusters with initial "s". If the syllable breaks fall right after the vowel of the penultimate of words such as 'em.ba.ssy, 'di.li.gent, 'si.ni.stral, then the stress on the antepenultimate is consistent with the QSR. But if the syllable structure is 'em.bas.sy, 'di.lig.ent, 'si.nis.tral, then these words do not conform to the QSR. The cases in question are distinguished in the column as various subcases of "maybe". There aren't too many of them, so you can ignore them, at least for starters.
Consistent with Latin or Greek etymon? This column encodes whether the headword might (in principle) have been borrowed from Latin or Greek rather than from French. For instance, any word ending in the morphemes -able or -ion are in principle consistent with a Latin eytmon, even if the OED has evidence that the actual historical source is French. The reason the possible etymology is important is that native speakers don't have access to the actual history, but only to the words themselves, together with any morphological structure they impose on them.
Otherwise rule-governed? Probably best omitted.
Other comments Same as in Assignment 6.
"Ante" offset parameter The number of years that is subtracted from an "ante" date to turn it into the computationally tractable Effective OED date. Feel free to play around with this parameter if you want.

Class
Corrections	Sort on this column to locate updates and corrections.
Investigator	Same as in Assignment 6.
Headword
Usage
OED date
Effective date	Here are the guidelines that I've used to map the dates in the "OED date" column into the computationally tractable form in the "Effective date" column. circa (for instance, c1349) A "circa" date is the best estimate that the OED editors can give regarding the date of first attestation. In reducing such a date to a number, I ignored the "circa." ante (for instance, a1250) "Ante" (Latin for 'before') means that the loanword is believed to have been first attested before the date given. The problem is that there is no indication of how many years before. In assigning "ante" dates to time periods, I ignored the "ante" and subtracted a fixed number of years from the remaining date. I chose 25 years (the "offset" in the last column of the header row of the spreadsheet). You are free to experiment with this parameter. date ranges (for instance, 1305-6) I took the arithmetic mean of the range and rounded to the next year. implied date ranges (for instance, 135. or 14..) I took the (rough) midpoint of the implied range. In other words, I treated 135. as 1355, and 14.. as 1450. If there was a known date earlier than the midpoint of the range, I used the known date. question mark (for instance, ?1349, ?a1100, ?c1475) I ignored any question marks and treated the remainder according to the preceding guidelines.
Etymon	Same as in Assignment 6.
Syllables	Same as in Assignment 6.
Chaucer stress, verse-medial	Stress is recorded using the same conventions as for the "OED stress" column, augmented by the various considerations laid out in Assignment 8 (especially in item 3).
Chaucer stress, in rhyme
Chaucer comments
Shakespeare stress
Shakespeare comments
OED strees	Same as in Assignment 6.
Prefix 1	Information concerning a headword's morphological structure. Useful for testing various hypotheses. For instance, did English speakers categorize (at least some) prefixes as on a par with Germanic unstressable prefixes? Another hypothesis: Words containing certain suffixes (notably, -ion, -Vble, -Vnce), might be borrowed from French or from Latin, and speakers wouldn't necessarily know which language was the source language. As the numbers of loanwords from Latin went up during the Renaissance, speakers might have recategorized loanwords from French with these suffixes as loanwords from Latin (even if there is incontrovertible philological evidence that they were originally borrowed from French) and treated them in special ways. For some suffixes, I distinguished two forms (for instance, "Vnce" and "Vnce 2"). The "2" means that the root of the headword is itself an English word. The absence of a "2" indicates that the root is a "cran" morpheme.
Prefix 2
Suffix 2
Suffix 1
Phrasal etymon?	Was the etymon of the word a French phrase (rather than a simple word)? Mostly for my convenience, but feel free to use if helpful.
Initial epenthesis?	Mostly for my convenience, but feel free to use if helpful.
Final syllabic C in etymon	Mostly for my convenience, but feel free to use if helpful.
Consistent with Quantity Stress Rule?	As you will see, it turns out that many French loanwords are exceptions to both the French and the Germanic stress rule. This column encodes whether the stress on a word is predictable from a third stress rule - what I will here call the Quantity Stress Rule and what is generally known as the Latin Stress Rule. According to this rule, word stress falls on the first syllable of a disyllabic word. In words with more than two syllables, stress falls on the penultimate if it is heavy, and else on the antepenultimate. It is usually clear whether the penultimate is heavy, but not always. In particular, it is not clear where the syllable break falls in certain consonant sequences, such as geminates, affricates, and clusters with initial "s". If the syllable breaks fall right after the vowel of the penultimate of words such as 'em.ba.ssy, 'di.li.gent, 'si.ni.stral, then the stress on the antepenultimate is consistent with the QSR. But if the syllable structure is 'em.bas.sy, 'di.lig.ent, 'si.nis.tral, then these words do not conform to the QSR. The cases in question are distinguished in the column as various subcases of "maybe". There aren't too many of them, so you can ignore them, at least for starters.
Consistent with Latin or Greek etymon?	This column encodes whether the headword might (in principle) have been borrowed from Latin or Greek rather than from French. For instance, any word ending in the morphemes -able or -ion are in principle consistent with a Latin eytmon, even if the OED has evidence that the actual historical source is French. The reason the possible etymology is important is that native speakers don't have access to the actual history, but only to the words themselves, together with any morphological structure they impose on them.
Otherwise rule-governed?	Probably best omitted.
Other comments	Same as in Assignment 6.
"Ante" offset parameter	The number of years that is subtracted from an "ante" date to turn it into the computationally tractable Effective OED date. Feel free to play around with this parameter if you want.