home > results and commentary > Trump vs. Biden (town halls)

Word Analysis of 2020 U.S. Presidential Debates

Donald Trump vs. Joe Biden (town halls)

15 October 2020



Introduction

The second debate was cancelled for reasons. Instead, simultaneous "duelling" town halls were held. I've used the transcripts from both and combined them to create a virtual debate transcript.

Both town halls were not of equal length. The moderator's closing remark were at 01:04:06 in Trump's town hall and at 1:32:02 in Biden's. So Biden's hall was

92 64
longer than Trump's. Keep this in mind when evaluating the total number of sections, sentences and words. Though, as we'll see, the rate of word delivery for both candidates was within 6% — a negligible difference.

Speaking Turns and Interruptions

Here, I look at the length of each turn of uninterrupted speech.

Table 1
length of sections in words
The number of uninterrupted deliveries (sections), mode/median/mean length of sections in words, and the shortest section length in words that composed 10%, 50% and 90% of the debate.
speaker sections section length debate contiguity (L10 L50 L90)
Donald Trump
201
201
1.0 15.0 36.0
1.00015.0000000036.000
17 89 161
17.00089.000161.000
Joe Biden
125
125
1.0 75.0 77.9
1.00075.0000000077.888
67 128 188
67.000128.000188.000

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 1
legend
a b c
51025

a — section length (mode), shortest section length in 10% of debate

b — section length (median), shortest section length in 50% of debate

c — section length (mean), shortest section length in 90% of debate

bar — proportion of a:b:c

Table 1
commentary

Despite the fact that Biden's town hall was

92 64
longer than Trump's, Trump delivered
201 125
more sections of uninterrupted speech. Trump's section medial length was
15 75
shorter than Biden's — a factor of 5× The means are a little closer, with Trump's mean section length
36 77.9
shorter than Biden. This indicates that Biden was more measured and paced in his delivery and did not have very short or very long sections to skew the mean away from the median.

Flesch-Kincaid Reading Ease and Grade Level

The Flesch-Kincaid reading ease and grade level metrics are designed to indicate how difficult a passage in English is to understand.

This metric does not take repetition into account. A grade level 10 sentence that is repeated 100 times still generates the same metrics because the words per sentence and syllables per word remain constant. To measure how many times a speaker repeats themselves, I use my Windbag Index, below.

Reading ease ranges from 100 (easiest) down to 0 (hardest) and can be interpreted as follows

100 –905th gradeVery easy to read. Easily understood by an average 11-year-old student.
90 – 806th gradeEasy to read. Conversational English for consumers.
80 – 707th gradeFairly easy to read.
70 – 608th & 9th gradePlain English. Easily understood by 13- to 15-year-old students.
60 – 5010th to 12th gradeFairly difficult to read.
50 – 30collegeDifficult to read.
30 – 10college graduateVery difficult to read. Best understood by college/university graduates.
10 – 0professionalExtremely difficult to read. Best understood by college/university graduates.

The grade level corresponds roughly to a U.S. grade level. It has a minimum value of –3.4 and no upper bound.

Two sets of readability scores are calculated. One for the entire debate and one that only considers section with at least 9 words.

Table 2a
readability — entire debate
Flesch-Kincaid reading ease and grade level.
speaker grade level reading ease sections sentences words syllables
Donald Trump
3.82
100.0%
3.82
84.57
100.0%
84.57
201
100.0%
201
764
100.0%
764
7,236
100.0%
7236
9,635
100.0%
9635
Joe Biden
6.41
167.8%
6.41
75.71
89.5%
75.71
125
62.2%
125
651
85.2%
651
9,736
134.5%
9736
13,343
138.5%
13343

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 2b
readability — excluding short sections
Flesch-Kincaid reading ease and grade level for sections with at least 9 words.
speaker grade level reading ease sections sentences words syllables
Donald Trump
4.16
100.0%
4.16
83.45
100.0%
83.45
138
100.0%
138
688
100.0%
688
7,025
100.0%
7025
9,385
100.0%
9385
Joe Biden
6.64
159.6%
6.64
75.04
89.9%
75.04
99
71.7%
99
623
90.6%
623
9,658
137.5%
9658
13,250
141.2%
13250

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 2
legend
a
b
30

a — value for candidate

b — value relative to Donald Trump

bar — proportion of a

Table 2
commentary

In the first debate, Trump's grade level was 3.60 and Biden's was 4.27. I made the argument that because of a larger number of interruptions by Trump, Biden's sentences were cut short of what they would have normally been, which pulled his grade level down.

The town halls allow a more fair comparison of the candidates' grade level in a debate-like format because unless the moderator stepped in, interruptions were rare.

Trump's grade level in the town hall is 3.82, which is

3.82 6.41
lower than Biden's. Interestingly, Trump's grade level in the town hall was almost identical to what it was in the first debate (
3.8 3.6
higher), suggesting that Biden had little influence over Trump's grade level.

Biden, on the other hand, had a 6.41 grade level, which was

6.41 4.27
higher than in the first debate.

There are at least two possibilities here: Trump lowers the grade level of his opponents to roughly his or trump lowers the grade level of his opponents by 2 grades.

Sentence Size

Table 3
sentence size
Number of sentences spoken by each speaker and sentence word count statistics. Number of words in a sentence is shown by average and 50%/90% cumulative values for all, stop and non-stop words.
speaker number of sentences sentence size
all stop non-stop
Donald Trump
765
765
9.5 13 29
9.52813.00029.000
5.7 8 19
5.6608.00019.000
3.9 5 12
3.8685.00012.000
Joe Biden
651
651
15.1 21 46
15.11421.00046.000
8.7 12 26
8.69612.00026.000
6.4 9 21
6.4189.00021.000
total
1,416
1416
14.1 18 40
14.09618.00040.000
9.1 11 23
9.05611.00023.000
7.0 8 18
7.0408.00018.000

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 3
legend
a b c
51025

a — average sentence size

b — largest sentence size for 50% of content

c — largest sentence size for 90% of content

bar — proportion of a:b:c

Table 3
commentary

Despite that Biden's hall was

92 64
longer than Trump's, Trump delivered
765 651
more sentences. Counting all words, Trump's sentences were on average
13 21
shorter than Biden's.

Coming soon.

Word Statistics

Debate Word Count

Summary Word Count

The summary word count reports the total number of words and the number of unique, non-stop words used by each candidate. Word number is expressed as both absolute and relative values.

Table 4a
all words
Number of all words and unique words used by each speaker.
set word count
Donald Trump
7,289 1,042
42.6% 14.3%
62471042
Joe Biden
9,839 1,629
57.4% 16.6%
82101629
total
17,128 2,120
100.0% 12.4%
150082120

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 4b
exclusive and shared words
Words exclusive to speaker (e.g. speaker A but not speaker B) and shared by speakers (speaker A and B).
set word count
Donald Trump
901 491
12.4% 54.5%
410491
Joe Biden
1,868 1,078
19.0% 57.7%
7901078
both candidates
14,359 551
83.8% 3.8%
13808551

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 4
legend
a c
b d
3010

a — word count

b — word count, as fraction in total in debate

c — unique words in (a)

d — unique words in (a), as fraction in (a)

bar — proportion of (a-c):c

Table 4
commentary

Biden delivered

9839 7289
more words than Trump, but part of this is due to the fact that Biden's Biden's town hall was
92 64
longer than Trump's. If we normalize for town hall length (64 min for Trump and 92 min for Biden), we find Trump delivered
114 107
more words per minute of town hall.

Biden delivered

1868 901
more unique words than Trump, a substantially larger number even if you account for the difference in total word count. Biden also had
1078 491
more exclusive (spoken by one candidate but not the other) words than Trump.

Stop Word Contribution

In the table below, the candidates' delivery is partitioned into stop and non-stop words. Stop words (full list) are frequently-used bridging words (e.g. pronouns and conjunctions) whose meaning depends entirely on context. The fraction of words that are stop words is one measure of the complexity of speech.

Table 5a
non-stop words
Counts of stop and non-stop words.
speaker all words stop words non-stop words
Donald Trump
7,289 1,042
100.0% 14.3%
62471042
4,330 134
59.4% 3.1%
4196134
2,959 908
40.6% 30.7%
2051908
Joe Biden
9,839 1,629
100.0% 16.6%
82101629
5,661 147
57.5% 2.6%
5514147
4,178 1,482
42.5% 35.5%
26961482
total
17,128 2,120
100.0% 12.4%
150082120
9,991 153
58.3% 1.5%
9838153
7,137 1,967
41.7% 27.6%
51701967

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 5b
exclusive and shared non-stop words
Non-stop words exclusive to speaker (e.g. speaker A but not speaker B) and shared by speakers (speaker A and B).
set word count
Donald Trump
890 485
30.1% 54.5%
405485
Joe Biden
1,832 1,059
43.8% 57.8%
7731059
both candidates
4,415 423
61.9% 9.6%
3992423

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 5
legend
a c
b d
3010

a — total number of words, for a given category (all, stop, non-stop)

b — (a) relative to words in the debate if category=all, otherwise relative to words by the candidate

c — number of unique words with set (a)

d — (c) relative to (a)

bar — proportion of (a-c):c

Table 5
commentary

Both candidates used roughly the same fraction of stop words, 59.4% for Trump and 57.5% for Biden. This is nearly the same fraction as in the 1st debate. Once stop words are removed, the spread of unique and exclusive words between Biden and Trump remains roughly the same — Biden had

1832 890
and
1059 485
more, respectively.

Word frequency

The word frequency table summarizes the frequency with which words were used. I show the average word frequency and the weighted cumulative frequencies at 50 and 90 percentile. The average word frequency indicates how many times, on average, a word is used. For a given fraction of the entire delivery, the weighted cumulative frequency indicates the largest word frequency within this fraction (details about weighted cumulative distribution).

Table 6a
word use frequency
Average and 50%/90% percentile word frequencies.
speaker word frequency
all stop non-stop
Donald Trump
7.0 33 245
6.99533.000245.000
32.3 97 251
32.31397.000251.000
3.3 6 56
3.2596.00056.000
Joe Biden
6.0 35 292
6.04035.000292.000
38.5 94 354
38.51094.000354.000
2.8 5 39
2.8195.00039.000
total
8.1 60 547
8.07960.000547.000
65.3 187 608
65.301187.000608.000
3.6 7 60
3.6287.00060.000

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 6b
exclusive and shared non-stop word use frequency
Average and 50%/90% cumulative percentile word frequencies. Non-stop words exclusive to speaker (e.g. speaker A but not speaker B) and shared by speakers (speaker A and B).
set word frequency
Donald Trump
1.83 2 6
1.8352.0006.000
Joe Biden
1.73 2 6
1.7302.0006.000
total
3.63 7 60
3.6287.00060.000

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 6
legend
a b c
51025

a — average word frequency

b — largest word frequency in 50% of content

c — largest word frequency in 90% of content

bar — proportion of a:b:c

Table 6
commentary

Trump tended to repeat his words slightly more than Biden.

All further word use statistics represent content that has been filtered for stop words, unless explicitly indicated.

Part of Speech Analysis

In this section, word frequency is broken down by their part of speech (POS). The four POS groups examined are nouns, verbs, adjectives and adverbs. Conjunctions and prepositions are not considered. The first category (n+v+adj+adv) is composed of all four POS groups.

Part of Speech Count

Table 7
part of speech count
Count of words categorized by part of speech (POS).
part of speech
n+v+adj+adv nouns (n) verbs (v) adjectives (adj) adverbs (adv)
Donald Trump
2,798 860
38.4% 30.7%
69044079725523114217766
1,130 440
40.4% 38.9%
690440
1,052 255
37.6% 24.2%
797255
373 142
13.3% 38.1%
231142
243 66
8.7% 27.2%
17766
Joe Biden
3,864 1,393
39.3% 36.1%
109477788843624520613682
1,871 777
48.4% 41.5%
1094777
1,324 436
34.3% 32.9%
888436
451 206
11.7% 45.7%
245206
218 82
5.6% 37.6%
13682
total
6,662 1,860
38.9% 27.9%
197410271809567529295341120
3,001 1,027
45.0% 34.2%
19741027
2,376 567
35.7% 23.9%
1809567
824 295
12.4% 35.8%
529295
461 120
6.9% 26.0%
341120

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 7
legend
a c
b d
1535

a — total number of words for a given POS (all, noun, verb, adjective, adverb)

b — (a) relative to all words by candidate

c — unique words in (a)

d — (c) relative to (a)

bar — proportion of (a-c):c

Table 7
commentary

We saw in the first debate that Trump used proportionatelly fewer nouns than Biden but he used more adverbs.

Here we see the same trend. Biden used Δrel=+19.8% (Δabs=+8.0%, 48.4% vs 40.4%) more nouns and Trump used Δrel=+55.4% (Δabs=+3.1%, 8.7% vs 5.6%) more adverbs. Trump used slightly more verbs and adjectives.

Part of Speech Frequency

Table 8
part of speech frequency
Frequency of words categorized by part of speech (POS).
part of speech frequency
n+v+adj+adv nouns (n) verbs (v) adjectives (adj) adverbs (adv) pronouns (pro)
Donald Trump
3.25 6 56
3.2536.00056.000
2.57 4 18
2.5684.00018.000
4.12 10 115
4.12510.000115.000
2.63 4 23
2.6274.00023.000
3.68 7 19
3.6827.00019.000
29.86 136 354
29.857136.000354.000
Joe Biden
2.77 5 40
2.7745.00040.000
2.41 4 24
2.4084.00024.000
3.04 6 139
3.0376.000139.000
2.19 3 12
2.1893.00012.000
2.66 5 39
2.6595.00039.000
29.27 162 254
29.266162.000254.000
total
3.58 7 60
3.5827.00060.000
2.92 5 35
2.9225.00035.000
4.19 12 254
4.19012.000254.000
2.79 4 34
2.7934.00034.000
3.84 9 53
3.8429.00053.000
51.38 298 608
51.377298.000608.000

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 8
legend
a b c
51025

a — average word frequency

b — largest word frequency in 50% of content

c — largest word frequency in 90% of content

bar — proportion of a:b:c

Table 8
commentary

Here we can see which parts of speech were more likely to be repeated by each candidate.

Just as in the first debate, Trump repeated his verbs more than any other part of speech:

10 6
more than Biden.

Part of Speech Pairing

Through word pairing, I extract concepts from the text. The number of unique word pairs is a function of sentence length and is one of the measures of complexity.

Table 9a
part of speech pairing — Donald Trump
Word pairs (total and unique) categorized by part of speech (POS)
part of speech pairings - Donald Trump
noun verb adjective adverb
noun
120 84
  70.0%
3684
verb
46 41
  89.1%
541
2 2
  100.0%
02
adjective
190 141
  74.2%
49141
0 0
  0.0%
00
6 5
  83.3%
15
adverb
2 2
  100.0%
02
45 37
  82.2%
837
21 14
  66.7%
714
10 5
  50.0%
55

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 9b
part of speech pairing — Joe Biden
Word pairs (total and unique) categorized by part of speech (POS)
part of speech pairings - Joe Biden
noun verb adjective adverb
noun
222 182
  82.0%
40182
verb
74 69
  93.2%
569
12 12
  100.0%
012
adjective
250 214
  85.6%
36214
6 5
  83.3%
15
14 10
  71.4%
410
adverb
5 5
  100.0%
05
32 29
  90.6%
329
9 9
  100.0%
09
3 3
  100.0%
03

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 9c
unique part of speech pairing — candidate comparison
Unique word pairs categorized by part of speech (POS)
unique part of speech pairings
noun (n) verb (v) adjective (adj) adverb (adv)
noun
84 182
  216.7%
84
182
verb
41 69
  168.3%
41
69
2 12
  600.0%
2
12
adjective
141 214
  151.8%
141
214
0 5
  0.0%
0
5
5 10
  200.0%
5
10
adverb
2 5
  250.0%
2
5
37 29
  78.4%
37
29
14 9
  64.3%
14
9
5 3
  60.0%
5
3

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 9 a,b
legend
a c
  d
3010

a — total number of pairs, for a given category (e.g. verb/noun)

c — number of unique pairs within set (a)

d — (c) relative to (a)

bar — proportion of (a–c):c

Table 9c
legend
a c
  d
50
45

a — unique pairs for Donald Trump

c — unique pairs for Joe Biden

d — (c) relative to (a) (i.e. Joe Biden relative to Donald Trump)

bars — (a) and (c)

Table 9
commentary

Trump had

45 32
more adverb/verb pairs and
21 9
more adverb/adjective pairs than Biden.

Look at that noun/noun pair difference though! Biden had

222 120
more such pairs. These included "N95 masks", "tax credit", "drug abuse" and "community policing" — all pairs exclusive to Biden. Trump had pairs like "campaign trail", "garbage cans", "travel ban" and "feeding frenzy" — none of these were used by Biden.

Detailed Part of Speech Tags

You can really get into the weeds here. Parts of speech are counted more granularly in these tables — nouns and verbs are split into classes and many other word types are shown, such as conjunctions and prepositions.

Table 10a
detailed POS tags — nouns and verbs
Count by part of speech tag: NN (noun, singular), NNP (proper noun, singular), NNPS (proper noun, plural), NNS (noun plural), VB (verb, base form), VBD (verb, past tense), VBG (verb, gerund/present participle), VBN (verb, past participle), VBP (verb, sing. present, non-3d), VBZ (verb, 3rd person sing. present)
Penn Treebank part of speech tag
NN NNP NNPS NNS VB VBD VBG VBN VBP VBZ
Donald Trump
691
9.48%
691
267
3.66%
267
11
0.15%
11
238
3.27%
238
396
5.43%
396
258
3.54%
258
155
2.13%
155
116
1.59%
116
520
7.13%
520
227
3.11%
227
Joe Biden
1,193
12.13%
172.6%
1193
318
3.23%
119.1%
318
16
0.16%
145.5%
16
394
4.00%
165.5%
394
571
5.80%
144.2%
571
337
3.43%
130.6%
337
255
2.59%
164.5%
255
165
1.68%
142.2%
165
495
5.03%
95.2%
495
319
3.24%
140.5%
319

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 10b
detailed POS tags — adjectives, pronouns, adverbs and wh-words
Count by part of speech tag: JJ (adjective), JJR (adjective, comparative), JJS (adjective, superlative), PRP (personal pronoun), PRP$ (possessive pronoun), RB (adverb), RBR (adverb, comparative), RBS (adverb, superlative), WDT (wh-determiner), WP (wh-pronoun), WP$ (possessive wh-pronoun), WRB (wh-abverb)
Penn Treebank part of speech tag
JJ JJR JJS PRP PRP$ RB RBR RBS WDT WP WP$ WRB
Donald Trump
373
5.12%
373
21
0.29%
21
15
0.21%
15
1,107
15.19%
1107
64
0.88%
64
568
7.79%
568
10
0.14%
10
3
0.04%
3
40
0.55%
40
53
0.73%
53
56
0.77%
56
Joe Biden
459
4.67%
123.1%
459
39
0.40%
185.7%
39
21
0.21%
140.0%
21
1,019
10.36%
92.1%
1019
104
1.06%
162.5%
104
578
5.87%
101.8%
578
9
0.09%
90.0%
9
3
0.03%
100.0%
3
51
0.52%
127.5%
51
124
1.26%
234.0%
124
87
0.88%
155.4%
87

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 10c
detailed POS tags — prepositions, conjunctions, determiners and others
Count by part of speech tag: CC (coordinating conjunction), CD (cardinal digit), DT (determiner), EX (existential there), FW (foreign word), IN (preposition/subordinating conjunction), MD (modal), PDT (predeterminer), POS (possessive ending), RP (particle), TO (to), UH (interjection)
Penn Treebank part of speech tag
CC CD DT EX FW IN MD PDT POS RP TO UH
Donald Trump
352
4.83%
352
74
1.02%
74
635
8.71%
635
13
0.18%
13
3
0.04%
3
620
8.51%
620
131
1.80%
131
6
0.08%
6
13
0.18%
13
42
0.58%
42
193
2.65%
193
18
0.25%
18
Joe Biden
392
3.98%
111.4%
392
206
2.09%
278.4%
206
973
9.89%
153.2%
973
49
0.50%
376.9%
49
1
0.01%
33.3%
1
1,034
10.51%
166.8%
1034
168
1.71%
128.2%
168
14
0.14%
233.3%
14
14
0.14%
107.7%
14
67
0.68%
159.5%
67
354
3.60%
183.4%
354
10
0.10%
55.6%
10

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 10
legend
a
b
c
10

a — total number of words with a given tag

b — (a) relative to all tagged words

c — (a) relative to number of words with this tag used by Donald Trump

bar — proportion of a

Table 10
commentary

This table is for data lovers. Let's look through the columns for any remarkable differences.

We've already seen that Biden delivered more nouns and this is reflected in the larger values in the NN* columns and that Trump used proportionately more adverbs RB*.

Trump had proportionately Δrel=+46.6% (Δabs=+4.8%, 15.19% vs 10.36%) more personal pronouns PRP, which I look at more detail below.

Exclusive and Shared Usage

This section enumerates words that were exclusive to a candidate (e.g. used by one candidate but not the other). This content provides insight into what the candidates' priorities are and reveals differences in perspective on similar topics.

For a given part of speech, the table breaks down the number of words that were spoken by only one of the candidates or both candidates (intersection). The last row includes words spoken by either candidate (union).

Table 11
exclusive word usage
Total and unique words used exclusively by a candidate, or by both.
part of speech
n+v+adj+adv nouns (n) verbs (v) adjectives (adj) adverbs (adv)
Donald Trump
864 467
100.0% 54.1%
13.0% 25.1%
397467
1872418712177823237
428 241
49.5% 56.3%
14.3% 23.5%
187241
187241
208 121
24.1% 58.2%
8.8% 21.3%
87121
87121
159 82
18.4% 51.6%
19.3% 27.8%
7782
7782
69 37
8.0% 53.6%
15.0% 30.8%
3237
3237
Joe Biden
1,732 1,000
100.0% 57.7%
26.0% 53.8%
7321000
404564163298871422450
968 564
55.9% 58.3%
32.3% 54.9%
404564
404564
461 298
26.6% 64.6%
19.4% 52.6%
163298
163298
229 142
13.2% 62.0%
27.8% 48.1%
87142
87142
74 50
4.3% 67.6%
16.1% 41.7%
2450
2450
both candidates
4,066 393
100.0% 9.7%
61.0% 21.1%
3673393
135519015331243395328228
1,545 190
38.0% 12.3%
51.5% 18.5%
1355190
1355190
1,657 124
40.8% 7.5%
69.7% 21.9%
1533124
1533124
392 53
9.6% 13.5%
47.6% 18.0%
33953
33953
310 28
7.6% 9.0%
67.2% 23.3%
28228
28228
total
6,662 1,860
100.0% 27.9%
100.0% 100.0%
48021860
197410271809567529295341120
3,001 1,027
45.0% 34.2%
100.0% 100.0%
19741027
19741027
2,376 567
35.7% 23.9%
100.0% 100.0%
1809567
1809567
824 295
12.4% 35.8%
100.0% 100.0%
529295
529295
461 120
6.9% 26.0%
100.0% 100.0%
341120
341120

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 11c
legend
a d
b e
c f
4030
40302015105

a — total number of words in set (e.g. obama \ romney, obama ∩ romney, obama ∪ romney , for a given part of speech

b — (a) relative to all exclusive words in n+v+adj+adv

c — (a) relative to all words in n+v+adj+adv

d — unique words in (a)

e — (d) relative to (a)

f — (d) relative to all unique words in n+v+adj+adv

bar1 — normalized ratio of (a-d):d

bar2 — absolute ratio of (a-d):d for all POS groups (first column) or POS group (other columns)

Table 11
commentary

Looking at each candidate's unique words that were exclusive to them, Biden had

546 241
more nouns,
289 121
more verbs,
142 82
more adjectives and
50 37
more adverbs. These are very large differences and reflect that the things that Biden said that Trump didn't say were more varied.

Pronoun Usage

This section explores pronoun use in detail. Refer to the methods section for details.

Pronoun Count

Fraction of all words that were pronouns.

Table 12a
pronoun fraction
Fraction of words that were pronouns.
speaker all pronouns
Donald Trump
7,289 1,042
100.0% 14.3%
62471042
1,672 56
22.9% 3.3%
161656
Joe Biden
9,839 1,629
100.0% 16.6%
82101629
1,873 64
19.0% 3.4%
180964
total
17,128 2,120
100.0% 12.4%
150082120
3,545 69
20.7% 1.9%
347669

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 12b
exclusive and shared pronouns
Pronouns exclusive to speaker (e.g. speaker A but not speaker B) and shared by speakers (speaker A and B).
set word count
Donald Trump
10 5
0.3% 50.0%
55
Joe Biden
20 13
0.6% 65.0%
713
both candidates
3,515 51
99.2% 1.5%
346451

Hover over fields with (e.g. 155) to download the corresponding data file.

Pronoun by Person, Gender and Count

Pronoun usage by person (1st, 2nd, 3rd), gender (masculine, feminine, neuter) and count (singular, plural).

Table 13a
Pronoun by person
Count of pronouns by first, second or third person.
pronoun person
all first second third
Donald Trump
1,172 20
100.0% 1.7%
5667188239811
573 7
48.9% 1.2%
5667
190 2
16.2% 1.1%
1882
409 11
34.9% 2.7%
39811
Joe Biden
1,124 22
100.0% 2.0%
5077233336212
514 7
45.7% 1.4%
5077
236 3
21.0% 1.3%
2333
374 12
33.3% 3.2%
36212

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 13b
Pronoun by gender
Count of pronouns by masculine, feminine or neuter gender.
pronoun gender
all masculine feminine neuter
Donald Trump
282 7
100.0% 2.5%
3333922032
36 3
12.8% 8.3%
333
41 2
14.5% 4.9%
392
205 2
72.7% 1.0%
2032
Joe Biden
252 8
100.0% 3.2%
693921663
72 3
28.6% 4.2%
693
11 2
4.4% 18.2%
92
169 3
67.1% 1.8%
1663

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 13c
Pronoun by number
Count of pronouns by singular or plural.
pronoun number
all singular plural
Donald Trump
1,339 41
100.0% 3.1%
9442535416
969 25
72.4% 2.6%
94425
370 16
27.6% 4.3%
35416
Joe Biden
1,337 44
100.0% 3.3%
8922740117
919 27
68.7% 2.9%
89227
418 17
31.3% 4.1%
40117

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 13
legend
a b
c d
153

a — total number of pronouns, by type

b — unique pronouns in (a)

c — (a) as fraction of all pronouns

d — (b) as fraction in (a)

bar — proportion of (a – b):b

Table 13
commentary

Proportionately, Trump used Δrel=+20.5% (Δabs=+3.9%, 22.9% vs 19%) more pronouns than Biden.

Looking at pronouns by person (1st, 2nd, 3rd), Trump and Biden used almost the same number of such pronouns. Trump used

573 514
and
409 374
more 1st and 3rd person pronouns while Biden used
236 190
more 2nd person pronouns.

Use of pronouns categorized by gender can be skewed by who is on the ticket. Given that Biden's ticket has a female, Trump's use of

41 11
more feminine pronouns is not surprising. Similarly, because both of Trump's ticket candidates are male, Biden's use of
72 36
more male pronouns is also not surprising.

First and third person pronouns — a closer look

These tables break pronouns by interesting contrasts. For example, the ratio of singular to plural 1st person pronouns reveals the use of "I/my/myself" vs. "we/our/ours".

Table 14a
1st person pronouns, by count
Count of singular and plural first person pronouns. This table contrasts use of I/my/myself vs. we/our/ours.
pronoun
first first singular first plural
Donald Trump
573 7
100.0% 1.2%
40041663
404 4
70.5% 1.0%
4004
169 3
29.5% 1.8%
1663
Joe Biden
514 7
100.0% 1.4%
31431934
317 3
61.7% 0.9%
3143
197 4
38.3% 2.0%
1934
Table 14b
3rd person pronouns, by count
Count of singular and plural third person pronouns. This table contrasts he/she/his/her/it vs. they/them/theirs.
pronoun
third third singular third plural
Donald Trump
409 11
100.0% 2.7%
27571234
282 7
68.9% 2.5%
2757
127 4
31.1% 3.1%
1234
Joe Biden
374 12
100.0% 3.2%
24481184
252 8
67.4% 3.2%
2448
122 4
32.6% 3.3%
1184
Table 14c
Me and you — 1st person singular and second person pronouns
Count of 1st person singular and second person pronouns. This table contrasts me/my/myself vs you/yours/yourself.
pronoun
all 1st singular 2nd
Donald Trump
594 6
100.0% 1.0%
40041882
404 4
68.0% 1.0%
4004
190 2
32.0% 1.1%
1882
Joe Biden
553 6
100.0% 1.1%
31432333
317 3
57.3% 0.9%
3143
236 3
42.7% 1.3%
2333
Table 14d
I, me, myself and my — closer look at 1st person singular pronouns
Count of specific 1st person singular pronouns: I, me, myself and my.
pronoun
all I me myself my
Donald Trump
403
100.0%
354.00035.0000.00014.000
354
87.8%
354.000
35
8.7%
35.000
0
0.0%
0.000
14
3.5%
14.000
Joe Biden
317
100.0%
254.00029.0000.00034.000
254
80.1%
254.000
29
9.1%
29.000
0
0.0%
0.000
34
10.7%
34.000
Table 14
legend
a b
c d
153

a — total number of pronouns, by type

b — unique pronouns in (a) (if more than one)

c — (a) as fraction of all pronouns

d — (b) as fraction in (a) (if less than 100%)

bar — proportion of (a – b):b

Table 14
commentary

Trump used

404 317
more 1st person singular pronouns, in keeping with his self-referencing style. Biden had a Δrel=+29.8% (Δabs=+8.8%, 38.3% vs 29.5%) proportion of 1st person plural pronouns of all his 1st person pronouns.

When contrasting the use of 1st person singular (me, myself) vs 2nd person (you), Biden had Δrel=+33.4% (Δabs=+10.7%, 42.7% vs 32%) proportionately more use of the 2nd person.

Pronouns by Category

This table tallies the use of pronoun by category. The categories are personal, demonstrative, indefinite, object, possessive, interrogative, others, relative, reflexive. Note that some pronouns that belong to multiple categories are counted in only one. For a list of pronouns for each category, see the pronoun methods section.

Table 15
Pronouns by cateogry
Count of pronouns by category.
pronoun category
all personal demonstrative indefinite object possessive interrogative others relative reflexive
Donald Trump
1,672
100.0%
1037.000182.000186.00078.00055.00067.00044.00022.0002.000
1,037
62.0%
10307
182
10.9%
1784
186
11.1%
16422
78
4.7%
735
55
3.3%
496
67
4.0%
634
44
2.6%
386
22
1.3%
211
2
0.1%
02
Joe Biden
1,873
100.0%
935.000245.000217.00078.000103.000140.000100.00049.0008.000
935
49.9%
9287
245
13.1%
2414
217
11.6%
19225
78
4.2%
735
103
5.5%
967
140
7.5%
1337
100
5.3%
946
49
2.6%
481
8
0.4%
53
Table 15
legend
a b
15

a — total number of pronouns, by category

b — (a) as fraction of all pronouns

bar — proportion of (a)

Table 15
commentary

As expected, most pronouns used were personal. Biden had slightly higher use of demonstrative pronouns.

Noun Phrase Usage

Noun phrases were extracted from the text and analyzed for frequency, word count, unique word count and richness. Single-word phrases were not counted.

Top-level noun phrases are those without a parent noun phrase (a parent phrase is one that a similar, longer phrase). Derived noun phrases are those with a parent (more details about noun phrase analysis).

The top-level noun phrases can be interpreted as independent concepts. Derived noun phrases can be interpreted as variants on concepts embodied by the top-level phrases.

Noun Phrase Count and length

This table reports the absolute number of noun phrases, which is related to the number of nouns, and their length.

Table 16a
noun phrase count
Counts of noun phrases in words and per noun.
speaker noun phrase count
all top-level
Donald Trump
314 158
100.0% 50.3%
0.00 0.00
156158
288 154
91.7% 53.5%
0.00 0.00
134154
Joe Biden
481 263
100.0% 54.7%
0.00 0.00
218263
462 258
96.0% 55.8%
0.00 0.00
204258

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 16b
noun phrase length
Average and 50%/90% cumulative length of noun phrases, in words.
speaker noun phrase length
all top-level
Donald Trump
2.11 2 3
2.1112.0003.000
2.12 2 3
2.1222.0003.000
Joe Biden
2.16 2 3
2.1642.0003.000
2.17 2 3
2.1692.0003.000

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 16a
legend
a d
b e
c f
1070

a — number of noun phrases

b — (a) relative to number of all noun phrases

c — number of noun phrases per noun

d — number of unique phrases

e — (c) relative to (a)

f — number of unique noun phrases per unique noun

bar — normalized ratio of (a–c):c

Table 16b
legend
a b c
102080

a — average noun phrase size, in words

b — largest noun phrase size in 50% of content

c — largest noun phrase size in 90% of content

bar — proportion of a:b:c


Table 16
commentary

Biden delivered

481 314
more noun phrases than Trump, a substantially larger number.

Exclusive and Shared Noun Phrase Count and length

Table 17a
exclusive and shared noun phrase count
Counts of exclusive and shared noun phrases in words and per noun.
speaker noun phrase count
all top-level
Donald Trump
277 141
34.8% 50.9%
136141
258 138
93.1% 53.5%
120138
Joe Biden
451 249
56.7% 55.2%
202249
436 247
96.7% 56.7%
189247
both candidates
67 17
8.4% 25.4%
5017
56 14
83.6% 25.0%
4214

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 17b
exclusive and shared noun phrase length
Average and 50%/90% cumulative length of noun phrases, in words.
speaker noun phrase length
all top-level
Donald Trump
2.13 2 3
2.1262.0003.000
2.14 2 3
2.1362.0003.000
Joe Biden
2.17 2 3
2.1752.0003.000
2.18 2 3
2.1792.0003.000
both candidates
2.00 2 2
2.0002.0002.000
2.00 2 2
2.0002.0002.000

Hover over fields with (e.g. 155) to download the corresponding data file.

Table 17a
legend
a c
b d
1070

a — number of noun phrases

b — (a) relative to number of all noun phrases

c — number of unique phrases

d — (c) relative to (a)

bar — normalized ratio of (a–c):c

Table 17b
legend
a b c
102080

a — average noun phrase size, in words

b — largest noun phrase size in 50% of content

c — largest noun phrase size in 90% of content

bar — proportion of a:b:c


Table 17
commentary

Biden had

451 227
more exclusive noun phrases than Trump.

Windbag Index

The Windbag Index is a compound measure that characterizes the complexity of speech. A low index is indicative of succinct speech with low degree of repetition and large number of independent concepts.

Unlike the Flesch-Kincaid readability metrics, the Windbag Index does not take into account the length of sentences or complexity (e.g. number of syllables) of individual words.

Table 18
windbag index
Windbag Index for each speaker. The higher the value, the more repetitive the speech.
speaker Windbag Index
index value index terms
Donald Trump
0
+0.0%
0
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
+0.0% +0.0% +0.0% +0.0% +0.0% +0.0% +0.0% +0.0%
<div>2959.000 7289.000</div><div>908 2959.000</div><div>440 1130.000</div><div>255 1052.000</div><div>142 373.000</div><div>66 243.000</div><div>158 314.000</div><div>154 158</div>
Joe Biden
0
+0.0%
0
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
+0.0% +0.0% +0.0% +0.0% +0.0% +0.0% +0.0% +0.0%
<div>4178.000 9839.000</div><div>1482 4178.000</div><div>777 1871.000</div><div>436 1324.000</div><div>206 451.000</div><div>82 218.000</div><div>263 481.000</div><div>258 263</div>
Table 18
legend
The Windbag Index is 1/(t1*t2*...*t9) where t1,t2,...,t8 are

t1 — fraction of words that are non-stop

t2 — fraction of non-stop words that are unique

t3 — fraction of nouns that are unique

t4 — fraction of verbs that are unique

t5 — fraction of adjectives that are unique

t6 — fraction of adverbs that are unique

t7 — fraction of noun phrases that are unique

t8 — fraction of noun phrases that are top-level


Large individual terms t1...t9 contribute to a smaller index.

The percentage values below the index and each term are relative differences to the other speaker's corresponding term (i.e. 100*(a-b)/b where a is the value for one speaker and b for the other).
Table 18
commentary

Despite the fact that Biden delivered

9839 7289
more words than Trump, Trump's Windbag Index was
1677 526
larger than Biden's. The total number of words tends to create a larger Windbag Index because the fraction of unique words drops.

In comparison to the 1st debate, Trump's Windbag index increased by

1677 1200
whereas Biden's dropped slightly by
526 539
.

Word Clouds

In the word clouds below, the size of the word is proportional to the number of times it was used by a candidate (method details).

Not all words from a group used to draw the cloud fit in the image — less frequently used words for large word groups may fall outside the image.

All Words for Each Candidate

Each candidate's debate portion was extracted and frequencies were compiled for each part of speech (noun, verb, adjective, adverb), with words colored by their part of speech category.

The distribution of sizes within a tag cloud follows the frequency distribution of words. However, word size cannot be compared between clouds, since the minimum and maximum size of the words is fixed.

Debate Word Cloud for Donald Trump - all words

Debate tag cloud for Donald Trump
Size proportional to word frequency. Color encodes part of speech: noun verb adjective adverb

Debate Word Cloud for Joe Biden - all words

Debate tag cloud for Joe Biden
Size proportional to word frequency. Color encodes part of speech: noun verb adjective adverb
commentary

Trump's main words were "great", "people", "know" and "never". Biden used "able" a lot.

Exclusive Words for Each Candidate

The clouds below show words used exlusively by a candidate. For example, if candidate A used the word "invest" (any number of times), but candidate B did not, then the word will appear in the exclusive word tag cloud for candidate A.

Words exclusive to Donald Trump

Debate tag cloud for Donald Trump
Size proportional to word frequency. Color encodes part of speech: noun verb adjective adverb

Words exclusive to Joe Biden

Debate tag cloud for Joe Biden
Size proportional to word frequency. Color encodes part of speech: noun verb adjective adverb
commentary

Trump said "maybe" a lot and it's surprising that Biden never said this word. Of course, "tremendous", "incredible" and "strong" are all part of Trump's lingo. Again, these words were never used by Biden.

Biden's exclusive words were "need", "school", "young" and "together".

Pronouns for Each Candidate

Word clouds based on only pronouns.

Pronouns for Donald Trump

Debate tag cloud for Donald Trump
Size proportional to word frequency. Color encodes pronoun type: masculine feminine neuter 1st person 2nd person singular plural other

Pronouns for Joe Biden

Debate tag cloud for Joe Biden
Size proportional to word frequency. Color encodes pronoun type: masculine feminine neuter 1st person 2nd person singular plural other
commentary

Pronoun usage was very similar.

Part of Speech Word Clouds

In these clouds, words from each major part of speech were colored based on whether they were exclusive to a candidate or shared by the candidates.

The size of the word is relative to the frequency for the candidate — word sizes between candidates should not be used to indicate difference in absolute frequency.

Cloud of noun words, by speaker

Words unique to each candidate (Trump, Biden) and those spoken by both.
commentary

In the 1st debate, we saw this cloud be full of blue words in the center (Biden repeated his exclusive words more often, proportionately).

Now, the situation is reversed and it's interesting to see which words exclusive to Trump got repeated a lot. These were "taxes", "decision", "antifa", "audit", "ballots" and "nancy".

Cloud of verb words, by speaker

Words unique to each candidate (Trump, Biden) and those spoken by both.
commentary

Trump said "saw" and "denounce" while Biden used "need" and "provide".

Cloud of adjective words, by speaker

Words unique to each candidate (Trump, Biden) and those spoken by both.
commentary

As expected, Trump's "tremendous" and "incredible" were common. Biden used "old", "young" and "significant".

Cloud of adverb words, by speaker

Words unique to each candidate (Trump, Biden) and those spoken by both.
commentary

Biden said "together" whereas Trump used "maybe". In the 1st debate Biden also used "together" frequently and exclusively. Two debates in and Trump has never said "together".

Cloud of all words, by speaker

Words unique to each candidate (Trump, Biden) and those spoken by both.
commentary

The center is nearly red except for "school", used by Biden.

Word Pair Clouds for Each Candidate

Pairs used only once during the debate are not shown.

word pairs for Donald Trump

JJ/JJ by Donald Trump
JJ/RB by Donald Trump
JJ/N by Donald Trump
JJ/V by Donald Trump
RB/RB by Donald Trump
RB/N by Donald Trump
RB/V by Donald Trump
N/N by Donald Trump
N/V by Donald Trump
V/V by Donald Trump

word pairs for Joe Biden

JJ/JJ by Joe Biden
JJ/RB by Joe Biden
JJ/N by Joe Biden
JJ/V by Joe Biden
RB/RB by Joe Biden
RB/N by Joe Biden
RB/V by Joe Biden
N/N by Joe Biden
N/V by Joe Biden
V/V by Joe Biden
commentary

Trump's pairs were "much better", "white supremacy", "fauci said" and "game changed". Biden's pairs were "new green", "long time", "fastest growing", "community policing".

Downloads

Debate transcript

Parsed word lists and word clouds (word lists, part of speech lists, noun phrases, sentences) (word clouds)

Raw data structure

Please see the methods section for details about these files.