Participial adjectives and the nouns that love them

At the request of Mr. Mark Brierley after the last post on participial adjectives, I’ve tried to come up with a few tests of my ambiguously defined “association” between participial adjectives and the verbs that they bear some family relation to. These tests probably have very little validity (bearing in mind that I never really defined “close” to begin with), but they do come in legitimate-looking tables.

All of them here are variations on what I call the “shared subject/object” test. Basically, if a participial adjective like “interesting” often has the same subject that a verb like “interest” does, or a participial adjective like “interested” often has the same subject that a verb like “interest” takes as an object, we can call them “close”.

I only did these searches for 5 verb/adjective sets, because until my institution pays for BYU corpora membership, I can only do so many searches a day. Also, this many searches on the iWeb corpus is already quite unhealthy.

So first, some raw numbers.

verb subjpres adj subj
#pron#n%pron#pron#n%pron
excite155193862.31%17773340583.92%
annoy306947686.57%888547694.92%
overwhelm70474248.69%3517351350.03%
inspire10137534765.47%95590151.45%
amaze474436292.91%955593532573.01%

These are the numbers of subjects of these verbs and their present participial adjectives which are pronouns and nouns respectively, followed by the percent of the total number of subjects that are pronouns. As you can see, the verb “annoy” comes with a pronoun subject (usually “it”) 86.57% of the time, while the adjective “annoying” has a pronoun subject even more often (also usually “it”). I didn’t keep track of how many of these are the dummy “it” seen in sentences like “it annoys me that robo-callers always spoof numbers from your own area code”, but I suspect it is a lot.

Next we have the differences between those rates of pronoun subjects:

excite21.61%
annoy8.34%
overwhelm1.34%
inspire14.01%
amaze19.90%

According to this probably unscientific measure, the verb “overwhelm” is “closer” to its present participial adjective “overwhelming” than “excite” is to “exciting”.

Let’s make this a little easier to visualize. On this test of “closeness”, the rankings are:

Test #1 ranking
excite5
annoy2
overwhelm1
inspire3
amaze4

Here is the same data for past participial adjectives. Because past participial adjectives are related to the passive voice of transitive verbs (e.g. “it excites me” is similar in meaning to “I am excited” rather than “it is excited”), I compared the rates of pronouns as the objects of verbs rather than subjects.

verb objpast adj subj
#pron#n%pron#pron#n%pron
excite840776091.71%68728591392.08%
annoy891056994.00%320548386.90%
overwhelm168424187.48%6292396.47%
inspire248861071369.91%83512487.07%
amaze1169823798.01%39188318292.49%
excite0.37%
annoy7.09%
overwhelm8.99%
inspire17.16%
amaze5.52%

This time, “excite” seems “closer” to “excited” than “overwhelm” is to “overwhelmed”, but there is a problem that invalidates most if not all of what I think I can learn from this particular group of searches. The searches that I used to get these results were overwhelms _nn*, overwhelms _p*, _nn* [be] overwhelmed_j* and _p* [be] overwhelmed_j*. Do you see the problem? If you do, you probably use the BYU corpora as much as I do (seek help!). When you search for _nn* [be] overwhelmed_j*, you will get back a lot of noun phrases that happen to end with a noun, for example “the man in the back row was overwhelmed“, but not so for overwhelms _nn*, which requires the first word after “overwhelms” to be a noun to return anything. English noun phrases being what they are, a lot of perfectly good noun phrases that start with articles or adjectives are not being detected by my search. That is, “overwhelms people” comes back as a hit, but “overwhelms a person” does not. In short, my searches for the relative frequencies of subjects and objects are not very valid – at best they tell me that pronouns are fairly common (pronoun searches do not suffer from the problem above since they don’t follow articles or adjectives). The same search problem vitiates all of my verb/past participial adjective searches, so take them all with grains of salt. I’m not going to give rankings to these due to the Sochi-like untrustworthiness of the results.

Now for something a bit more concrete: What specifically are the nouns that come with these verbs and adjectives?

verb subj #1#2#3pres adj subj #1#2#3
excitelightworkenergyfuturelifehome
annoythingnoisevoicethingnoisevoice
overwhelmworldlovelifeevidenceresponseodds
inspirestoryworkmusicstorypassionwork
amazeloveplacethingfoodproductstuff

Here we see some interesting things – the top 3 subjects for both verbs and present participial adjectives are the same for “annoy”/”annoying”, and 2 of the three are the same for “inspire”/”inspiring”. Here is an update ranking of “closeness”:

Test #1 rankingTest #2 ranking
excite53 (tie)
annoy21
overwhelm13 (tie)
inspire32
amaze43 (tie)

verb obj #1#2#3past adj subj #1#2#3
excitepeople electronsstudentsteampeoplecompany
annoypeople usersotherspeoplefansothers
overwhelmpeople studentsvisitorsheartsystemclinics
inspirepeople confidenceotherspostporkowl-ways
amazepeople visitorsaudiencespeopledoctorskids

First of all, the #1 object for all of the verbs is “people”, as in “it excites people” or “it inspires people”. There is more variety in the subjects of past participial adjectives, but 2 of the top hits are still “people” – “people are annoyed” and “people are amazed”. The objects suffer from the same validity problem outlined earlier, so take this with even more salt, but we can say at least that “people” is a common object for verbs whose past participial adjective counterparts have the subject “people”.

Last but not least, the #1 hits as a percent of the total hits. The first column is the #1 subject for the verb as a % of the hits for that verb, and the second column is the #1 subject for that adjective as % of the hits for that verb. The 3rd column is the 1st divided by the 2nd – the closer to 1 that number is, the more similar the proportion of those two words in the hits.

verb subj #1 as % of totalpres part adj #1 as % of totalratiopres adj subjverb subj as % of totalratio
excite0.0460.0143.2950.0370.00162.500
annoy0.1030.1031.0000.0880.0881.000
overwhelm0.0420.0049.6670.2190.00635.000
inspire0.0570.0590.9680.1590.1591.000
amaze0.0300.0083.6670.0710.00327.648

Based on the above, “annoy” and “annoying” are quite “close”, but “overwhelm” and “overwhelming” are not. This is the last test to count toward official rankings:

Test #1 rankingTest #2 rankingTest #3 ranking
excite53 (tie)4
annoy211
overwhelm13 (tie)5
inspire322
amaze43 (tie)3

That gives us an idea of which of the verb/present participial adjective pairs is “closest” according to all of the tests I ran. But first, the same results for verbs/past participial adjectives:

verb obj #1 as % of total past part adj #1 as % of total #1 as % of total past adj subjverb obj #1 as % of totaltotal nouns
excite0.1600.00360.3770.1600.0632.534
annoy0.4650.4651.0000.2960.2961.000
overwhelm0.2860.000Div 00.1300.0433.000
inspire0.1490.000Div 00.0410.000Div 0
amaze0.2570.2571.0000.2260.2261.000

Likewise for verbs and past participial adjectives – “annoy” and “inspire” are closer to “annoyed” and “inspired” than “overwhelm” is to “overwhelmed”.

So to close this post, here are the final overall rankings for “closeness” of verbs and present participial adjectives:

Test #1 rankingTest #2 rankingTest #3 rankingOverall
excite53 (tie)45
annoy2111
overwhelm13 (tie)53
inspire3222
amaze43 (tie)34

“Annoy” is the clear winner, followed not too distantly by “inspire”.

Since I still have the data that I used in the last post, I can tell you that these rankings correlate with the % of uses that were verbs at -0.51 (that is, numerically higher rankings, i.e. 4 and 5, tended to be used as verbs less than as adjectives). That says to me that the words that have a fully established independent life as adjectives tend to have less in common collocation-wise with their verb parents. This makes sense – the average child hears the adjective “exciting” applied to all kinds of activities before he ever learns that light “excites” electrons and has to form the relationship between those two words post hoc. I guess, although I don’t know, that “inspire” and “inspiring” might tend to be acquired much closer to the same time in a child’s life and in more similar situations.

As always, further research is required! Who has a paid BYU corpora membership?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s