The classic
“To be honest…”
Usually means: I am about to admit something, disagree, or give you the version I want you to treat as especially candid.
The irritation: it can sound as though honesty has just been switched on.
A fact page · with Beatrix
A small inquiry into the expressions people use when they want us to know that what comes next is honest, frank, serious, or—most suspiciously—“not gonna lie”.
Read01 / The family finding
Beatrix and I discovered that we share a particular irritation: the expression “to be honest”. Not because honesty is a bad thing. Quite the opposite. It is the announcement that bothers us.
If the sentence is honest, why does it need a little honesty siren fitted to the front? And if it does need one, what are we supposed to infer about the sentences that arrived without it?
This page is not a case against ordinary speech. It is an attempt to notice what these tiny phrases are doing.
02 / The phrases
In linguistics, many of these are called discourse markers or pragmatic markers: expressions that help manage a conversation and signal attitude.
The classic
Usually means: I am about to admit something, disagree, or give you the version I want you to treat as especially candid.
The irritation: it can sound as though honesty has just been switched on.
The internet cousin
Often used before an embarrassing admission, a criticism softened by friendliness, or a surprisingly strong opinion.
The joke: it denies a lie that nobody has accused anyone of telling.
The family-size version
This may be the expression Beatrix heard. It means “I am telling you this is true”, and it belongs to the same family of truth-emphasising phrases.
If the phrase was “honest to God”, the mechanism is similar—just with a larger witness.
It wasn’t me.
And anyway,
it was an accident.
03 / The accidental defence
There is a wonderfully compressed piece of human reasoning here. First: I deny responsibility. Second: I offer a reason that would matter only if I were responsible.
It is not a coherent defence so much as a conversational escape hatch. The speaker may be trying to reduce blame, end the interrogation, or reach for the first explanation that sounds socially acceptable. The words come out in a familiar pattern, so the contradiction often passes unnoticed.
That is the interesting part: ordinary speech is full of prefabricated moves. We can say a sentence fluently before we have checked whether its parts agree.
04 / What the phrases do
The words may add little factual information, but they can still change the social temperature of a sentence.
“To be honest” tells the listener how to receive what follows: as a confession, correction, or personal judgement.
“Not gonna lie” can make a criticism or complaint feel less formal and less confrontational.
A familiar phrase gives the speaker a beat to organise the next thought while keeping the turn in conversation.
Sometimes the phrase is less a report about truth and more a public display of the speaker’s desired relationship with truth.
05 / A provisional verdict
Perhaps that is what bothers us: the phrase asks for credit before the evidence arrives. But it also reveals something wonderfully human. We are not only exchanging information. We are constantly trying to manage trust, embarrassment, conflict and the impression we make while doing it.
Still, if you say “to be honest” one more time, Beatrix and I reserve the right to ask what you were doing in the previous sentence.
Back to End Up Here06 / Add to the collection
Send us the expression that makes you stop listening—or the strange bit of everyday logic you cannot believe people say out loud.
07 / The growing collection
Approved contributions will appear here with a short research note where useful. Nothing is published automatically.
The collection is waiting for its first approved contribution.
08 / Sources and further reading
This page draws on the Cambridge Dictionary entry for “to be honest”, its entry for “not gonna lie”, Cambridge Grammar’s overview of discourse markers, and Fraser’s foundational paper “What are discourse markers?”. The family conversation and the “accident” example are supplied observations, not research findings.