Exploring Emptiness 2.Small prefixes 3. Semantic profiles 4. Constructional profiles 5. Prefix variation 6. Aspectual triplets Русская версия

Semantic Profiling

Chapter 3 "Big Prefixes: Semantic Profiling" aims to prove that the prefixes retain their meanings when they form Natural Perfectives. Chapter 2 analyzed the eleven “small” prefixes using the radial category profiling method. In this chapter we analyze the remaining five “big” prefixes, po-, s-, za-, na-, and pro-, using the semantic profiling method. Whereas the big prefixes pose a challenge for radial category profiling because they combine with a large number of verbs, this abundance is actually an advantage for semantic profiling. Semantic profiling requires data in larger quantities than are available for the small prefixes. Thus the two methods are complementary. Semantic profiling is analysis of the distribution of semantic tags. The Russian National Corpus (RNC) contains semantic tags that classify verbs according to their meanings. We thus have access to an independent classification of the meanings of verbs, and it is possible to see how these classifications are distributed across the prefixes used to form Natural Perfectives.

Statistical analysis

This website gives you access to the data and the statistical analyses that are reported in Chapter 3. This way you can both inspect the data and run the analyses yourself on your own computer. We will guide you through all the steps and provide some commentary, but if you are interested in learning more about using statistical analysis of linguistic data, we recommend that you consult the following books:

How to download R

You can download the R statistical software package to your computer from the R project webpage.

How to download and run the files from this website

On this webpage we offer you files that you can download with the ".R" extension. These are called "R scripts". You can download them by right-clicking on the links on this page. It is important that you download them to your home directory in your computer so that R can find them. If you do not know where your home directory is, you can also copy and paste the R commands from the scripts directly into the R window (see "Alternative methods for running R scripts" below). The R scripts contain all the commands that R needs in order to run the statistical tests. You can open the scripts if you like and see all the commands. We provide commentary on each command in lines that begin with the "#" symbol (R itself ignores all these lines) in order to help you follow along.

How to run the files from this website in R

After you have downloaded an R script, you can open the R program on your computer. At the ">" prompt, type in: source("") and put the name of the R script you want to run between the quotation marks. For example, you can enter a line that looks like this: > source("Ch3Table1.R") and when you hit the return key, R will run the R script and give you all of the results as output.

Alternative methods for running R scripts

If you simply click on the links with the R scripts, you can then copy and paste all of the code into the R window and R will run the commands and give you the same results. Another option is to download the R script to any location in your computer you want to and provide the path to the file when you use the source command. For example, you can enter a line that looks like this: > source("/Users/janedoe/Downloads/Ch3Table1.R") for Mac users or > source("C://Documents/Ch3Table1.R") for PC users. If you do not know the path, you can open your finder to where the R script is and then drag and drop that file into an open R window placing it after the cursor prompt ">". When you do this, R will tell you what the path to the file is and you can copy and paste that into the source command.

Evaluation of Table 1: Chi-square test and effect size

Dowload this R script to your home directory: Ch3Table1.R and run it according to the directions above. It will give you the chi-squared value, the degrees of freedom, the p-value, and the effect size. You can open and read the commentary in the R script to see how it is done.

CORRECTION: In the original version, there was an error in the calculation of the Cramer’s V value. The actual value is somewhat lower, but still indicates a large effect size. Here is an R script that gives the correct value: Ch3Table1Correct.R.

Expected values for Table 1

We used a simple formula involving the row and column totals in order to calculate the expected values for all the cells in Table 1. Here it is:
expected value = (row sum x column sum) / total sum
The total sum for Table 1 is 382, and all of the row and column totals are listed in Table 1 in Chapter 3. Download this R script to your home directory: Ch3Table1.exp.values.R and run it according to the directions above. It will compute and print out for you all of the expected values for Table 1 in Chapter 3.

Fisher Test values in Table 2

The Fisher Test values reported in Table 2 in Chapter 3 were calculated using this Fisher's Exact Test Calculator. For the cells in Table 1 where the expected value was less than the observed value, we calculated a right-sided p-value. These results are reported with [+] in Table 2 and indicate the probability that we would get this many items or more in the cell given the overall distribution of items. This includes, for example, the cells for pro-/‘sound’ & ‘speech’, po-/‘changest’, etc. For the cells in Table 1 where the expected value was greater than the observed value, we calculated a left-sided p-value. These results are reported with [-] in Table 2 and indicate the probability that we would get this many items or fewer in the cell given the overall distribution of items. This includes, for example, the cells for za-/‘sound’ & ‘speech’, po-/‘impact’, etc. The Fisher test calculates that chance of getting the same or an even more extreme value given the overall distribution of values in a matrix. In order to compute the Fisher test probability, you need to put in four values. These values relate the value in the cell to the sum for the row, the sum for the column, and the sum for the entire table. Here is a sample of the values that would be used for computing the Fisher test probability for pro-/‘sound’ & ‘speech’:

a = value in the given cell
= 51
b = row total - value in the given cell
= 106 - 51
= 55
c = column total - value in the given cell
= 65 - 51
= 14
d = table total - value in the given cell - b - c
= 382 - 51 - 55 - 14
= 262

When we insert these values into the Fisher test calculator on the website and click on compute, it returns the probabilities. Because a = 51 is greater than the expected value (which is only 18 according to the calculations above), we want to know the chances that a could be this large or larger given the overall distribution, so we calculate a right-sided value. The calculator lists this as “prob a >= 51” and gives us the value 2.946e-21. This is the probability that we could get 51 or more verbs in this cell, and is the same as the value that we report in Table 2 in Chapter 3. Below are all of the values needed to use the calculator for the combinations of prefixes and semantic tags, following the cells in Table 1 and indicating which probability (prob a <= x or prob a >= x) should be used, according to the relative sizes of the actual and expected values. You can use these to compute the Fisher test probabilities found in Table 2 of Chapter 3.

CORRECTION: In the original version, there was an error in the calculation of d in this table. This resulted in the incorrect calculation of Fisher test p-value. The table below gives the correct d and the next table shows the correct Fisher test p-values.

semantic tag po- s- na- za- pro-
‘impact’ a = 11, b = 111
c = 110, d = 150
prob a <= 11
a = 23, b = 99
c = 43, d = 217
prob a >= 23
a = 31, b = 91
c = 28, d = 232
prob a >= 31
a = 47, b = 75
c = 24, d = 236
prob a >= 47
a = 10, b = 112
c = 55, d = 205
prob a <= 10
‘changest’ a = 62, b = 40
c = 59, d = 221
prob a >= 62
a = 11, b = 91
c = 55, d = 225
prob a <= 11
a = 3, b = 99
c = 56, d = 224
prob a <= 3
a = 22, b = 80
c = 49, d = 231
prob a >= 22
a = 4, b = 98
c = 61, d = 219
prob a <= 4
‘behav’ a = 11, b = 41
c = 110, d = 220
prob a <= 11
a = 23, b = 29
c = 43, d = 287
prob a >= 23
a = 17, b = 35
c = 42, d = 288
prob a >= 17
a = 1, b = 51
c = 70, d = 260
prob a <= 1
a = 0, b = 52
c = 65, d = 265
prob a <= 0
‘sound’ &‘speech’ a = 37, b = 69
c = 84, d = 192
prob a >= 37
a = 9, b = 97
c = 57, d = 219
prob a <= 9
a = 8, b = 98
c = 51, d = 225
prob a <= 8
a = 1, b = 105
c = 70, d = 206
prob a <= 1
a = 51, b = 55
c = 14, d = 262
prob a >= 51

semantic tag po- s- na- za- pro-
‘impact’[-] 4e-12[+] 0.33[+] 0.0003[+] 3.7e-11 [-] 0.0009
‘changest’[+] 6.3e-13[-] 0.03[-] 5.8e-06[+] 0.22[-] 6.3e-06
‘behav’[-] 0.052[+] 6.5e-07[+] 0.0006[-] 0.0001[-] 2.8e-05
‘sound’ &‘speech’[+] 0.23[-] 0.003[-] 0.005[-] 1.9e-10[+] 3e-21

Full list of verbs in this study

There are 382 verbs in this study, and a full list of these verbs, arranged according to the five prefixes and the semantic tags, is shown on this page. The table below shows what semantic tags are attracted to, neutral to or repulsed from each prefix. The listings in the table work as a hyperlink: if you click on it, you will be directed to the full list of verbs representing this class.

RUS / ENG page

Each page exists in two versions: Version RUS provides all the Russian data using the Cyrillic alphabet, while Version ENG provides the same data transliterated into the Latin alphabet. Note that the lists of verbs are organized in Russian alphabetic order.

prefix po- s- na- za- pro-
Attractions 62 ‘changest’ verbs 23 ‘behav’ verbs
23 ‘impact’ verbs
31 ‘impact’ verbs
17 ‘behav’ verbs
47 ‘impact’ verbs 51 ‘sound’ & ‘speech’ verbs
Neutral 37 ‘sound’ & ‘speech’ verbs
11 ‘behav’
- - 22 ‘changest verbs’ -
Repulsions 11 ‘impact’ verbs 9 ‘speech’ verbs
11 ‘changest’ verbs
3 ‘changest’ verbs
8 ‘speech’ verbs
1 ‘behav’ verb
1 ‘speech’ verb
4 ‘changest’ verbs
10 ‘impact’ verbs

Attractions: 62 ‘changest’ verbs
LemmaGloss
pobagrovet’ turn purple
pobelet’ turn white
poblednet’ turn pale
pobleknut’ lose color
poburet’ turn brown
povzroslet’ mature
povlažnet’ become damp
poglupet’ become stupid
pogolubet’ turn blue
podeševet’ become cheaper
podlinnet’ become longer
podorožat’ increase in price
podurnet’ grow ugly
poželtet’ turn yellow
pozdorovet’ turn recuperate
pozelenet’ turn green
pozelenit’ paint green
pozlatit’sja turn golden
pozolotet’ turn golden
pokrasnet’ turn red
pokrupnet’ grow larger
polevet’ shift leftward
polegčat’ lessen, abate
polilovet’ turn violet
pomolodet’ get younger-looking
pomračnet’ turn dark
pomutnet’ muddy
pomjagčet’ get soft
poplotnet’ grow stout; become dense
popolnet’ grow stout
popravet’ shift rightward
poprostet’ become unpretentious
poredet’ thin out
porozovet’ turn pink
poryžet’ turn reddish
porjabet’ get speckled
posaxarit’ sugar
posvežet’ freshen
posedet’ turn gray
poseret’ turn grizzly
poser’eznet’ become serious
poseč’sja tear
posivet’ turn gray
posizet’ turn blue-gray
posinet’ turn dark blue
posmirnet’ become quiet
posmuglet’ turn swarthy
pospet’ ripen
posurovet’ become severe
poteplet’ get warmer
potončat’ become thin
potreskat’sja split
potusknet’ turn dim
potusknut’ turn dim
poumnet’ grow wiser
poxolodet’ cool down
poxolodnet’ cool down
poxorošet’ grow prettier
poxrabret’ become braver
poxudat’ get slimmer
poxudet’ get slimmer
počernet’ turn black

Neutral: 37 ‘sound’ & ‘speech’ verbs (35 ‘speech’ and 2 ‘sound’)
LemmaGloss
poblagodarit’ thank
pobožit’sja swear
povinit’sja confess
povorožit’ tell fortunes
pogutorit’ chat
požalit’sja complain
požalobit’sja complain
požalovat’sja complain
pozvat’ call
pozvonit’sja ring
pozdorovat’sja say hello
pointeresovat’sja be curious about
pokljast’sja swear
polajat’sja bark
pomolit’sja pray
poobeščat’ promise
poobeščat’sja promise
popenjat’ blame
poprivetstvovat’ welcome
poprosit’ request
poprosit’sja ask for
poručit’sja guarantee
posetovat’ complain
posmejat’sja laugh
posovetovat’sja get advice
posporit’ argue
posulit’ promise
posulit’sja promise
potoropit’ hurry
potrebovat’ require
poxvalit’ praise
poxvalit’sja boast
poxvastat’ brag
poxvastat’sja brag
poxodatajstvovat’ intercede
pocapat’sja bicker
pošutit’ make a joke

Neutral: 11 ‘behav’
LemmaGloss
podelikatničat’ be soft on smby.
požadničat’ act greedy
pokuražit’sja swagger
polenit’sja act lazy
poostorožničat’ be over-cautious
poskupit’sja scrimp
poslušat’sja obey
postesnjat’sja be shy
potešit’sja amuse
poxlopotat’ hustle about
poceremonit’sja stand on ceremony

Repulsions: 11 ‘impact’ verbs
LemmaGloss
pobrit’ shave
pobrit’sja shave
podoit’ milk
podrat’sja fight
pokolotit’ wallop
pokorobit’ warp
pokryt’ cover
pokusat’ bite
počesat’sja scratch oneself
poščekotat’ tickle
poščupat’ touch, feel

Attractions: 23 ‘behav’ verbs
LemmaGloss
svelikodušničat’ do something generous
sglupit’ do something stupid
sdvurušničat’ be a double-dealer
s”exidničat’ speak maliciously
sžulit’ cheat
sžul’ničat’ cheat
slevačit’ work on the side
sliberal’ničat’ play the liberal
slovčit’ dodge
slukavit’ be cunning
smalodušestvovat’ act like a coward (once)
smalodušničat’ act like a coward (once)
smošenničat’ cheat
smuxlevat’ swindle
snaxal’ničat’ be impudent
sobez’janničat’ ape
sozorničat’ play tricks
soriginal’ničat’ do something original
spodličat’ act meanly
sprovocirovat’ provoke
sxalturit’ do something carelessly
sximičit’ contrive
sxitrit’ do something clever

Attractions: 23 ‘impact’ verbs
LemmaGloss
sbrošjurovat’ stitch up
svarganit’ knock off
svarit’ cook
svit’ twist; weave
svjazat’ tie, knit
sdelat’ make, do
sžat’ reap
sžeč’ burn down
skomkat’ crumple
skopnit’ stook
skulemat’ crumple
smasterit’ craft, build
smontirovat’ mount
sostrjapat’ cook
srabotat’ work out
sravnjat’ even out
srubit’ chop down
ssučit’ twist
strenožit’ hobble
sfabrikovat’ fabricate
sfal’cevat’ fold
sfugovat’ joint
sšit’ sew

Repulsions: 9 ‘speech’ verbs
LemmaGloss
sagitirovat’ campaign for
skalamburit’ make a pun
skomandovat’ command
sorientirovat’ orient
sostrit’ make a witty remark
srifmovat’ make a rhyme
storgovat’sja strike a bargain
sformulirovat’ formulate
s”jazvit’ say sth sarcastic

Repulsions: 11 ‘changest’ verbs
LemmaGloss
svarit’sja get cooked
sgnit’ rot
sgnoit’ let rot
sžat’sja shrink
skondensirovat’ condense
skosobočit’sja get lopsided
smjagčit’ make softer
sostarit’ age
sostarit’sja age
stvorožit’ curdle
stvorožit’sja curdle

Attractions: 31 ‘impact’ verbs
LemmaGloss
nabelit’sja whiten
nabronzirovat’ bronze
navaksit’ polish (shoe)
navorsit’ tease (fibers)
navorsovat’ tease (fibers)
navoščit’ wax
nagljancevat’ glaze
nagofrirovat’ crimp
nakanifolit’ apply rosin
nakostyljat’ give a good drubbing
nakraxmalit’ starch
namalevat’sja paint (ones face)
namelit’ chalk
namylit’ soap
namylit’sja soap (oneself)
naodekolonit’ sprinkle with cologne
naoxrit’ apply ochre
napičkat’ force-feed
naploit’ quill
napomadit’ pomade
napomadit’sja pomade oneself
napryskat’ sprinkle
nasandalit’ apply sandalwood
nasur’mit’ cover with surma
nasur’mit’sja paint onself with dye
nafabrit’ stain (mustaches)
nafabrit’sja stain (mustaches)
naxoxlit’ ruffle up
našinkovat’ shred
našpigovat’ apply lard
namozolit’ work until one has callouses

Attractions: 17 ‘behav’ verbs
LemmaGloss
nabedokurit’ get into mischief
nabezobrazničat’ behave disgracefully
nabuzit’ make a row
nagrubijanit’ be rude
nadeboširit’ kick up a row
naerundit’ play the fool
nakaverzit’ make tricks
nakaverzničat’ make tricks
nakurolesit’ play tricks
naplevat’ spit
naprokazit’ play pranks
naprokazničat’ play pranks
napyžit’sja puff up
nasvinjačit’ make a dirty act
naxuliganit’ behave like a hooligan

Repulsions: 8 ‘speech’ verbs
LemmaGloss
nagrubit’ be rude
naderzit’ be fresh with smb.
naklevetat’ gossip
nakljauzničat’ libel
naproročit’ prophesize
naspletničat’ gossip
naxamit’ be rude
najabedničat’ snitch

Repulsions: 3 ‘changest’ verbs
LemmaGloss
nadymit’ fill with smoke
naxolodit’ cool
naèlektrizovat’ electrify

Attractions: 47 ‘impact’ verbs
LemmaGloss
zaasfal’tirovat’ cover with asphalt
zaballastirovat’ load with ballast
zabarrikadirovat’ barricade
zabarrikadirovat’sja barricade oneself
zabetonirovat’ concrete
zabintovat’ bandage
zabodat’ gore
zabronirovat’ reserve
zabutit’ fill with rubble
zavualirovat’ veil
zagatit’ make a brushwood road
zagipsovat’ cover with gypsum
zagudronirovat’ cover road with tar
zadekorirovat’ decorate
zadrapirovat’ drape
zadrapirovat’sja drape
zazubrit’sja get notched
zakamuflirovat’ camouflage
zakonservirovat’ preserve
zakrepit’ fasten
zakristallizovat’ crystallize
zalatat’ patch
zamarkirovat’ mark
zamaskirovat’ mask
zamaskirovat’sja mask
zamesit’ knead
zaminirovat’ mine
zamordovat’ victimize
zamusolit’ slobber all over
zamyzgat’ make filthy
zaplatat’ patch
zaprudit’ dam up
zapjatnat’ soil
zasilosovat’ store in a silo
zaskirdovat’ mow up
zasuslit’ slobber all over
zasusolit’ slobber all over, suck
zatavrit’ brand
zatorcevat’ trim
zatuševat’ shade
zatjukovat’ pack in bales
zaxloroformirovat’ chloroform
zašnurovat’sja lace up
zašpaklevat’ putty
zašpuntovat’ groove
zaštopat’ darn
zaštukovat’ mend

Neutral: 22 ‘changest verbs’
LemmaGloss
zavšivet’ become lousy
zagrjaznit’ make dirty
zagrjaznit’sja get dirty
zagustet’ thicken
zaindevet’ get covered with frost
zakonspirirovat’ make secret
zakosnet’ stagnate
zalubenet’ harden
zamarinovat’ marinate
zamerznut’ freeze
zamuslit’sja get covered with slobber
zamusolit’sja get covered with slobber
zaplesnevet’ become moldy
zapylit’sja get covered with dust
zarubcevat’sja get covered with a scar
zastudenet’ become jellied
zatverdet’ harden
zatravenet’ get covered with grass
zaxilet’ become sickly
zaxiret’ fall into decay
začaxnut’ wither
zašeršavet’ become rough

Repulsions: 1 ‘behav’ verb
LemmaGloss
zadurit’ become unreasonable

Repulsions: 1 ‘speech’ verb
LemmaGloss
zakontraktovat’ make a contract

Attractions: 51 ‘sound’ & ‘speech’ verbs (23 ‘sound’ and 28 ‘speech’)
LemmaGloss
probasit’ talk in a deep voice
probormotat’ mutter
probrenčat’  jingle
probrexat’  yelp, bark
probrjuzžat’ grumble
probrjacat’ clank
probubnit’ mutter
proburčat’ grumble
provereščat’ squeal
proveščat’ prophesy
progavkat’ bark
progogotat’ cackle
progremet’ thunder
progromyxat’ rumble
progroxotat’ crash, rumble
progudet’ buzz
prodeklamirovat’ recite
prodiktovat’ dictate
prozvenet’ ring
prozvučat’ sound, be heard
proinstruktirovat’ instruct
prointerv’juirovat’ interview
proinformirovat’ inform
prokvakat’ croak
prokommentirovat’ comment
prokonsul’tirovat’ offer consultation
prokonsul’tirovat’sja consult
prokričat’ yell
prokukarekat’ crow
prokukovat’ cuckoo
prolajat’ bark
prolepetat’ babble
prolopotat’ splutter
promurlykat’ purr
promyčat’ moo
promjamlit’ hum and haw
promjaukat’ meow
propesočit’ criticize
prosvistet’ whistle
prosignalit’ signal
prosklonjat’ decline; bandy about
prosprjagat’ conjugate
protelegrafirovat’ telegraph
protelefonirovat’ telephone
protjavkat’ yelp
prourčat’ rumble
proxripet’ wheeze
procitirovat’ quote
prošeptat’ whisper
proèkzamenovat’ examine
proèkzamenovat’sja be examined

Repulsions: 10 ‘impact’ verbs
LemmaGloss
proburavit’ bore, drill
proburit’ bore, drill
proventilirovat’ ventilate
provejat’ winnow
prodezinficirovat’ apply disinfectant
prodiskovat’ break up soil with a disk-shaped harrow
proizvestkovat’ apply lime (soil)
promotyžit’ hoe
prosverlit’ drill, perforate
protaranit’ ram

Repulsions: 4 ‘changest’ verbs, 0 ‘behav’ verbs
LemmaGloss
progorknut’ become bitter
propitat’sja become saturated
proxudit’sja get worn out
projasnet’ clear up