What Hope Means to the British Public: The Lexical Appendix
DDI United Kingdom · Audience response study
Part 2 · AppendixThe Vocabulary of the Reply
The words the public used in reply to Prime Minister Andy Burnham’s “What Hope Means” video, and how they change by platform, reception profile and DDI band.
Part 2 of Let’s Bring Back Hope is one study published in three parts. All three draw on the same corpus of scored comments, and they are meant to be read together.
- ReportWhat Hope Means to the PublicThe analytical report: findings, interpretation, and implications for policy and leadership.
- Interactive mapThe Discourse MapEvery scored comment on the map, with a guided reading, a comparison tool and seminar exercises.
- Lexical appendixThe Vocabulary of the Reply You are hereThe words the public used, by platform, reception profile and DDI band.
Part 1: “Let’s Bring Back Hope” in 48 Hours, with its interactive discourse map and lexical appendix.
Seven readings
- Hope is still the first word, and it is still being returned. Hope appears in 2,668 comments, more than any other word. Its most frequent companions are the slogan’s own words, “bring” and “back”. Yet the next most common hope phrase is “no hope” (351 comments), and “hope is” (291) and “only hope” (145) follow close behind.
- The reply speaks the language of procedure and territory. After the slogan, the most frequent phrases are “general election” (617), “stop boats” (279) and “call general” (191).
- Each platform has its own register. Facebook is over-represented for boats, illegals, Reform and the election; X for masked profanity and the vocabulary of crime; TikTok for bills, prices, taxes, data centres and the working class; Instagram for climate, Gaza and genocide.
- Each reception profile has its own grammar. Hostile comments criminalise and dehumanise (“illegals”, “invasion”, “puppet”, “clown”, “traitor”). Dismissive comments talk about talk (“holiday”, “microphone”, “words”, “waffle”). Counter-demand comments specify the remedy (“election”, “call”, “vote”, “Reform”). Civil scepticism prices it (“tax”, “mortgage”, “prices”, “action”). Constructive comments locate it (“community”, “services”, “health”, “families”, “local”).
- The frame entered the vocabulary. “Holiday” ranks among the forty most used words. 349 comments mention the holiday or the recess, 78 the “How to Fail” microphone and 50 the music.
- Repetition is slogan, not campaign. The most repeated text is “general election now” (15 copies). There is no equivalent of Part 1’s single-issue “genocide genocide” pattern; genocide appears in 110 comments.
- Healthy discourse is still particular. The words most over-represented in Mixed and Healthy comments are “community”, “mental”, “health”, “local”, “services”, “families” and “poverty”. The language of the better reply is the language of a particular situation.
Counts are numbers of distinct comments containing a word, so a word repeated within one comment counts once. Exact repeats of a whole comment are counted once in the lexical analysis and examined separately in Section 6.
The corpus as a whole
Word clouds show salience, not meaning, and this appendix reads them alongside ranked counts, phrases and concordance lines. Across 11,895 distinct comments the reply uses 13,227 different content words. The most frequent are hope, people, back, country, stop, Labour and election.
The vocabulary of the reply
The 90 content words used in the most comments, all platforms. Size shows the number of comments using the word; words are set in alphabetical order.
action actually andy away back believe benefits better boats bring bringing british burnham call care change children cost country debt election enough feel first general getting give giving gone good government great hard help holiday hope illegal illegals immigrants immigration job labour last left life listen live living making man manchester migrants minister money needs never new nothing party pay paying people power prime public reform rid right sort starmer start still stop support system talk talking tax taxes time vote weeks well without words work working world year years
The thirty most used words
Number of distinct comments containing each word. The figure on the right is the rate per 1,000 comments.
Show the top 60
| Word | Comments | Per 1,000 |
|---|---|---|
| hope | 2,668 | 224.3 |
| people | 1,441 | 121.1 |
| back | 1,222 | 102.7 |
| country | 1,190 | 100.0 |
| stop | 1,024 | 86.1 |
| labour | 942 | 79.2 |
| election | 835 | 70.2 |
| bring | 782 | 65.7 |
| andy | 724 | 60.9 |
| general | 643 | 54.1 |
| boats | 545 | 45.8 |
| government | 516 | 43.4 |
| give | 503 | 42.3 |
| burnham | 498 | 41.9 |
| tax | 483 | 40.6 |
| money | 457 | 38.4 |
| reform | 408 | 34.3 |
| british | 407 | 34.2 |
| nothing | 395 | 33.2 |
| call | 374 | 31.4 |
| illegal | 361 | 30.3 |
| time | 361 | 30.3 |
| party | 351 | 29.5 |
| better | 350 | 29.4 |
| never | 348 | 29.3 |
| starmer | 337 | 28.3 |
| years | 319 | 26.8 |
| holiday | 306 | 25.7 |
| good | 299 | 25.1 |
| work | 299 | 25.1 |
| working | 294 | 24.7 |
| well | 293 | 24.6 |
| start | 290 | 24.4 |
| change | 285 | 24.0 |
| man | 285 | 24.0 |
| pay | 285 | 24.0 |
| right | 279 | 23.5 |
| job | 262 | 22.0 |
| vote | 256 | 21.5 |
| talk | 253 | 21.3 |
| words | 222 | 18.7 |
| still | 221 | 18.6 |
| taxes | 215 | 18.1 |
| living | 214 | 18.0 |
| first | 205 | 17.2 |
| public | 200 | 16.8 |
| immigrants | 200 | 16.8 |
| life | 197 | 16.6 |
| giving | 194 | 16.3 |
| action | 191 | 16.1 |
| rid | 191 | 16.1 |
| help | 190 | 16.0 |
| cost | 186 | 15.6 |
| prime | 182 | 15.3 |
| left | 182 | 15.3 |
| bringing | 181 | 15.2 |
| talking | 179 | 15.0 |
| great | 179 | 15.0 |
| getting | 178 | 15.0 |
| making | 178 | 15.0 |
The top of the list is dominated by the slogan and its answer. “Bring” and “back” rank in the top ten because the Prime Minister’s phrase was quoted back at him in hundreds of comments. “Stop”, “boats”, “illegal” and “immigrants” mark the borders argument; “election”, “general”, “call” and “vote” mark the demand for consent; “tax”, “money”, “pay” and “cost” mark the household ledger. “Holiday” sits among the forty most used words, which tells its own story.
Phrases that recur
Two-word sequences after stop words are removed, by number of comments. “Back hope” is what remains of “bring back hope”; “call general” of “call a general election”.
Hope in context
The word hope appears in 2,668 comments. Read in its phrases, it sorts into three families. There is the slogan, quoted back (“bring back hope”, “bringing back hope”). There is denial (“no hope”, “there is no hope”, “no hope with…”). And there is definition and condition (“hope is…”, “the only hope…”, “hope for a general election”).
Hope in its sentences
Phrases containing the word hope, with function words kept, by number of comments.
The words that gather around hope confirm the pattern. Beyond the slogan’s own verbs, the strongest associations are with loss (“lost”, “gone”, “zero”), with removal and procedure (“resign”, “calling”, “general”, “election”), and with a small vocabulary of possibility (“possibility”, “feeling”, “sense”, “future”). Hope is talked about as something taken away, something conditional on an election, and occasionally as something that might still be felt.
The company hope keeps
Words that appear within five words of hope more often than their overall frequency would predict. The ratio compares observed with expected comments.
Hope in the line
Concordance lines for four recurring constructions. Each line is a different comment.
“only hope” 147 comments
“no hope” 351 comments
“hope is / hope would be” 342 comments
“bring back hope” 358 comments
By platform
The four platforms share a core vocabulary but differ sharply in what they over-use. Facebook carries the borders and party argument: boats, illegals, migrants, Reform, Farage and the election. X is the most profane, and the most likely to reach for the vocabulary of crime: “rape”, “gangs”, “deportations”, “fraud”. TikTok talks about the household and the economy: bills, prices, taxes, data centres, the working class. Instagram’s reply is shaped by Gaza, genocide and climate.
Four platforms, four vocabularies
For each platform: the words most over-represented compared with the other three (weighted log-odds keyness), and its most frequent words.
6,240 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back better boats bring british burnham call change country election general give good government holiday hope illegal immigrants job labour man money never nothing party pay people reform right starmer start stop tax time vote well work working years
TikTok
3,141 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back better bring british burnham call change country election feel general give good government hope job labour life living man money never nothing pay people reform right start stop talk tax taxes time vote well words work working years
585 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back better bring burnham care change children climate country data free gaza genocide give good government help hope job labour life money needs never new north people public right start stop support system tax time water well work years
X
1,929 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back better boats bring bringing british burnham call country debt election f*** f***ing general give government holiday hope illegal immigration job labour man money never nothing party people s*** starmer start still stop talk tax time weeks word years
X’s profile is the lexical counterpart of its DDI score. The platform with the lowest composite in the Part 2 report is also the one whose most distinctive words are masked profanity. Just below the fourteen words shown, Facebook’s distinctive list includes the mocking nickname “Andy Pandy”, a reminder that much of the hostility is directed at the person rather than the policy.
By reception profile
The Part 2 report grouped comments into five reception profiles on the basis of their DDI indicator scores alone. The vocabulary was not used to build the profiles, which makes the clarity of their word lists informative.
Five reception profiles, five vocabularies
For each reception profile from the Part 2 report: its most distinctive and most frequent words.
Hostile
4,156 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back benefits better boats bring british burnham call clown country election general give government holiday hope illegal illegals immigrants labour man migrants money never nothing party pay people reform rid starmer start still stop tax time work working world
Dismissive
3,139 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back better boats bring british burnham change country give giving gone good government holiday hope job labour man money never nothing party pay people reform starmer still stop talk talking tax taxes time weeks well words work working years
Counter-demand
1,442 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back better boats bring british burnham call change country election first general give good government hope illegal immigrants immigration job labour money never party people public reform rid right start stop tax time vote voted well work working years
Civil scepticism
2,046 distinct comments
Most distinctive words keyness · comments
Most frequent words
action andy back better boats bring british burnham change cost country election first give good government holiday hope job labour living money never nothing pay people reform right start stop talk tax taxes time well words work working year years
Constructive
1,085 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back better bring burnham care change country election feel general give good government help hope labour life living making money needs new pay people prime public reform right start stop support system tax taxes time well work working years
Each profile has a grammar. Hostile comments criminalise and dehumanise: illegals, invasion, rapists, corrupt, puppet, clown, traitor. Dismissive comments talk about talk and about the messenger’s absence: holiday, microphone, words, waffle, weeks. Counter-demand comments identify a remedy: election, call, vote, Reform, the electorate. Civil scepticism prices the promise: tax, mortgage, prices, allowance, action, economy. Constructive comments locate it in a place and a service: community, services, health, mental, staff, local, families.
Read in sequence, the five lists describe a movement from the enemy, to the performance, to the procedure, to the price, to the place.
By DDI band
The same pattern appears when comments are grouped by DDI band. Risk-band comments over-use the borders and party vocabulary and the masked insult. Concerning comments turn to policy: pensions, VAT, duty, businesses, defence and water. Mixed and Healthy comments use the language of community, health and family.
Where score meets vocabulary
For each DDI band: its most distinctive and most frequent words. Mixed and Healthy are combined because only 20 comments reach Healthy.
Risk
9,475 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back better boats bring british burnham call country election general give good government holiday hope illegal immigrants job labour man money never nothing party pay people reform rid starmer start stop talk tax time well words work working years
Concerning
2,054 distinct comments
Most distinctive words keyness · comments
Most frequent words
action andy back better bring british burnham call change cost country election general give good government help hope job labour life living man money never pay people public reform right start stop tax taxes time vote well work working years
Mixed or Healthy
366 distinct comments
Most distinctive words keyness · comments
Most frequent words
andy back better bring burnham care change community country families feel future give good government health help hope labour life living local making money needs new people prime public right social stop support system tax time well work working years
The frame, the copy and the emoji
A notable share of the vocabulary concerns the message’s circumstances rather than its content. “Holiday” is the clearest case. Comments about the bus fare score above the corpus mean, because they tend to engage with a policy; comments about the music, the microphone and the holiday score below it.
Words about the video itself
Comments mentioning the circumstances, setting or production of the message, with the mean DDI composite of those comments.
Exact repetition is modest. Most repeated texts are short slogans that many people could type independently: “general election now”, “bring back a general election”, “actions speak louder than words”, “totally agree”. Two longer texts were posted several times: a list of government achievements on TikTok, and a hostile message about protesters on Facebook. Pseudonymous author links show that each came from a single account, posted nine and ten times respectively, within minutes.
Words said more than once
Texts that appear word for word in four or more comments, after handles, names and punctuation are removed.
| Text | Copies | Platforms | Mean DDI |
|---|---|---|---|
| “general election now” | 15 | TikTok 7, Facebook 6, X 2 | 18.6 |
| “totally agree” | 11 | Facebook 11 | 23.9 |
| “bring back a general election” | 10 | TikTok 5, Facebook 4, X 1 | 19.0 |
| A long message calling protesters “non patriotic” and accusing Reform of “brainwashing” them (text not reproduced) | 8 | Facebook 8 | 15.9 |
| “actions speak louder than words” | 8 | Facebook 4, TikTok 4 | 17.9 |
| “well said” | 7 | Facebook 7 | 29.7 |
| “call a general election” | 6 | Facebook 4, TikTok 2 | 18.5 |
| “stop the boats” | 6 | X 4, Facebook 2 | 10.8 |
| “reform desperate posts” | 5 | Facebook 5 | 2.2 |
| “general election please” | 5 | TikTok 3, Facebook 2 | 19.0 |
| A long list of government achievements beginning “New Border Force agency, immigration down…” | 5 | TikTok 5 | 27.2 |
| “change the record” | 4 | Facebook 3, Instagram 1 | 11.0 |
| “bring on a general election” | 4 | Facebook 3, TikTok 1 | 19.0 |
| “reform is our only hope” | 4 | Facebook 2, TikTok 2 | 30.8 |
The emoji of the reply
The twelve emoji used in the most comments. 15.0% of comments contain at least one.
| Platform | Comments with emoji | Most used |
|---|---|---|
| 15.7% | 😂 😡 🤣 🤡 💩 🇬🇧 | |
| TikTok | 15.5% | 😂 👍 🙏 😡 🙄 😳 |
| 19.9% | 😂 🙏 🇬🇧 🇮🇱 🔥 ❤ | |
| X | 10.3% | 🤣 🤔 😡 🤡 😂 🇬🇧 |
Laughter is the most common emoji, followed by anger. The Union flag and the clown face follow; TikTok leans toward thumbs-up and folded hands, and Instagram includes the Israeli flag, used mostly in comments about Britain’s relationship with Israel.
From Part 1 to Part 2
The two corpora differ in size, window and stop list, so the comparison is made on ranks. The core is stable: hope and people hold the first two places, and back, country and Labour remain in the top ten. What moved is telling. “Stop” rises from 12th to 6th, carried by the borders vocabulary. “Let”, the first word of “let’s bring back hope”, falls from 10th to 29th as the slogan of the first message gives way to the question of the second. “Tax” falls from 11th to 21st and “right” from 15th to 44th.
Part 1’s fifteen most used words, ranked again in Part 2
Rank in the Part 1 lexical appendix (July 2026) and rank of the same word in Part 2, on a comparable stop list.
Show the data
| Word | Part 1 rank | Part 2 rank | Part 2 comments |
|---|---|---|---|
| hope | 1 | 1 | 2,668 |
| people | 2 | 2 | 1,441 |
| will | 3 | 5 | 1,119 |
| back | 4 | 3 | 1,222 |
| country | 5 | 4 | 1,190 |
| election | 6 | 9 | 835 |
| andy | 7 | 11 | 724 |
| labour | 8 | 8 | 942 |
| now | 9 | 16 | 581 |
| let | 10 | 29 | 354 |
| tax | 11 | 21 | 483 |
| stop | 12 | 6 | 1,024 |
| need | 13 | 14 | 593 |
| general | 14 | 13 | 643 |
| right | 15 | 44 | 279 |
New words enter the upper ranks: “bring” (10th), “boats” (17th), “government” (18th), “Reform” (23rd), “illegal” (26th) and “holiday” (35th). Part 1’s single-issue “genocide” vocabulary has receded: the word appears in 110 Part 2 comments and ranks 148th.
Method
Corpus. The 12,139 comments in the Part 2 primary analysis cohort, across Facebook, TikTok, Instagram and X. Comments whose cleaned text was identical to an earlier comment (244) were counted once, leaving 11,895 distinct comments. Frame vocabulary, repetition and emoji use all comments.
Cleaning. Account handles and links were removed. Facebook replies begin with the name of the person being answered; these name tags (1,156 comments) were removed, and the resulting name tokens excluded from every count. Text was lower-cased; a possessive “’s” was dropped. Standard English stop words were removed, together with contractions and conversational fillers; “back”, “never”, “nothing”, “people” and “country” were kept because they carry meaning in this corpus. Words shorter than three letters were dropped. Words were not stemmed, so “illegal” and “illegals” are counted separately.
Counting. Every count is the number of distinct comments containing a word or phrase. Counts are of the exact word, so “hope” excludes “hopeless” and “hopefully”; the Part 2 report’s figure of 2,796 comments mentioning hope includes those forms and counts repeated comments. Phrases are sequences after stop words are removed, so “call general” stands for “call a general election”. The hope phrases in Section 2 keep function words.
Keyness. Distinctive words are identified with the weighted log-odds ratio with an informative Dirichlet prior (Monroe, Colaresi and Quinn, 2008), comparing each group with all others at the level of comments, and reported as a z-score. Words must appear in at least ten comments in the group (eight for bands, six for the smallest groups).
Collocation. Words within five positions of hope are counted once per comment and compared with the number expected from their overall frequency.
Groupings. Platforms, reception profiles and DDI bands are those of the Part 2 report. Profiles were derived from DDI indicator scores, not from vocabulary.
Display. Profanity is masked with asterisks. Slurs are not displayed. Concordance lines are drawn at random from comments without a machine-flagged allegation or a slur, and are trimmed around the phrase. The longer of the repeated texts are described rather than reproduced.
Limits. Word counts show what is said often, not what is meant; irony, quotation and sarcasm are invisible to them. Dictionary choices shape every list. The corpus is a captured, non-probability sample of the public reply, and the comparison with Part 1 is a comparison of ranks across two different designs.
Read the whole study
- ReportWhat Hope Means to the PublicThe analytical report: findings, interpretation, and implications for policy and leadership.
- Interactive mapThe Discourse MapEvery scored comment on the map, with a guided reading, a comparison tool and seminar exercises.
- Lexical appendixThe Vocabulary of the Reply You are hereThe words the public used, by platform, reception profile and DDI band.
References
- GCRD (2026). Let’s Bring Back Hope, Part 2: What Hope Means to the Public. GCRD Insights, October 2026.
- GCRD (2026). The Discourse Map: interactive companion to “Let’s Bring Back Hope, Part 2”. GCRD Insights, October 2026.
- GCRD (2026). The Vocabulary of the Reply: lexical appendix to “Let’s Bring Back Hope, Part 2”. GCRD Insights, October 2026.
- GCRD (2026). The vocabulary of the reply: lexical appendix to “Let’s Bring Back Hope” (Part 1). GCRD Insights, July 2026.
- GCRD (2026). “Let’s Bring Back Hope” in 48 Hours. GCRD Insights, July 2026.
- Monroe, B. L., Colaresi, M. P. and Quinn, K. M. (2008). Fightin’ words: lexical feature selection and evaluation for identifying the content of political conflict. Political Analysis, 16(4), 372–403.
Suggested citation. Jacob, J. U. (2026). The Vocabulary of the Reply: Lexical Appendix to Let’s Bring Back Hope, Part 2. Global Centre for Rehumanising Democracy, October 2026.
This is an independent analysis, prepared without commission from any party, campaign or government body. More on the Democracy Discourse Index at gcrd.org.uk/ddi.
Global Centre for Rehumanising Democracy · gcrd.org.uk · info@gcrd.org.uk
