Ontology, attribution, and the people hidden inside our sentences.

“They told me not to.”

A short sentence can introduce an entire audience into a relationship. Suddenly, two people are speaking in the presence of friends, relatives, advisers, or critics who are absent from the room. Their apparent judgment may become a reason to act, a source of fear, or an explanation for a decision already made.

But who has actually entered the conversation?

One person? Three people who offered different opinions? A group that reached a decision? Or people the speaker imagines would disapprove?

Our short video begins with that ordinary relational moment. This article develops the questions underneath it. The aim is a method we can use both when listening to others and when examining our own speech.

1. A pronoun points; it does not prove

Philosophers of language study expressions whose reference depends on context. “I” ordinarily identifies the speaker; expressions such as “they” may depend on an earlier mention, a gesture, or shared contextual understanding. The same word can identify different people on different occasions. This is normal linguistic functioning, not a defect in language.[1]

The practical difficulty arises when a listener cannot recover the intended reference, or when that reference changes during an argument.

Consider: “They are against us.” The sentence leaves several matters open. Who belongs to “they”? Who belongs to “us”? What does “against” mean here? Disagreeing with a proposal, declining an invitation, criticizing conduct, and planning harm are different claims.

Clarifying one word will not settle all of those questions. Identifying the people does not establish what they believe; establishing a belief does not establish coordinated action; and establishing an action does not settle whether it was justified.

We therefore separate four inquiries:

InquiryQuestion
ReferenceWhich people does the expression identify?
OntologyWhat kind of thing are we describing: individuals, a category, an association, or an organized decision-making body?
Evidence and attributionWhat happened, how do we know, and whose conduct or belief is being described?
AuthorityWhat permits someone to speak or decide for the others?

These questions overlap, but they are not interchangeable. A precise list of names can still support an inaccurate allegation. A legitimate organization can still be misrepresented by an unauthorized spokesperson.

2. Unpacking “they” into a set

We can use the idea of a set as a simple analytic aid. Ask what rule includes someone in the collection being discussed. “My friends” is one collection. “My friends who attended dinner” is smaller. “The friends who expressly advised me to cancel” may be smaller still.

Suppose five friends attended dinner. Two advised cancellation, one disagreed, and two said nothing. “They all told me to cancel” would misdescribe that example. “Two friends advised me to cancel” preserves the actual scope of the report.

Notice that this correction does not decide whether cancellation is wise. Two friends might offer excellent reasons. Five might offer poor ones. Counting supporters and assessing their reasons are separate tasks.

The method is to move from the label to the relevant people, then from those people to their actual words or conduct. We need not turn every conversation into a roster. A sufficiently clear description—“two friends who were at dinner”—may do the necessary work without identifying anyone publicly.

Nor does the existence of a set establish a collective actor. Five people can share a characteristic without sharing an intention. Giving them one label does not supply a common mind.

3. The “ghost” in the conversation

We use “ghost” here as a metaphor for an apparent social presence whose basis remains unclear. It is not a diagnosis, and it does not mean that the underlying people or danger must be imaginary.

“They will judge me” might mean “two relatives have criticized this before.” It might mean “I expect my colleagues to disapprove.” It might mean “I feel ashamed and imagine a wider audience sharing that judgment.” Those are different accounts, each deserving a different response.

The feeling can be real while the account of its source remains uncertain. We can acknowledge distress and still ask what supports the prediction. Equally, uncertainty does not require us to dismiss a credible concern until every person involved has been identified.

What clarification offers is a more proportionate object of attention. We can discuss a relative’s actual criticism, a prediction about colleagues, or the speaker’s own conflict. An indefinite audience otherwise risks becoming larger and more unanimous than the evidence warrants.

When “they” remains undefined, the appropriate conclusion may be “we do not yet know who is meant”—not “nobody exists,” and not “the allegation is established.”

A refusal to clarify may justify withholding confidence from a claim. It does not, by itself, establish deliberate dishonesty. The speaker may be protecting someone, struggling to remember, or unable to explain clearly. The claim and the explanation for its vagueness require separate assessment.

4. “We decided”: membership, agreement, and agency

Philosophical work on collective intentionality examines what makes intending or acting together different from merely having similar individual intentions. There are competing accounts of that relationship; shared intention is not adequately explained simply by placing several people under one description.[2]

In a relationship, “we decided” can conceal several possible histories. Both people expressly agreed. One proposed something and interpreted silence as agreement. One made a choice within authority previously delegated by the other. Or one chose alone and subsequently described the choice as joint.

Those histories matter. The relevant inquiry is not just who belongs to the couple. It is how this particular decision became attributable to both people, if it did.

Sometimes collective decision-making does not require unanimity: a committee may use an agreed voting rule. A couple may authorize one partner to handle a routine matter. But a decision procedure or delegation needs a basis; the pronoun alone supplies neither.

It is also possible to belong to a meaningful “we” without endorsing every action taken within it. Shared identity, shared affection, shared intention, and authority to represent another person should not be treated as synonyms.

A pronoun can report agreement. It cannot substitute for the agreement it reports.

For relationship agreements, the practical application is concrete: record what has actually been settled, what remains a proposal, and where either person may act independently. Clear language can protect both the relationship and the individuality of its participants.

5. “I” brings the inquiry back to the speaker

“I” may be comparatively easy to identify, but the claim following it still needs examination.

“I heard two friends criticize the plan” reports an observation. “I think their criticism is right” expresses an assessment. “I am worried about disappointing them” describes a concern. “I have decided not to go” owns a decision.

These statements might all be true. Compressing them into “they won’t let me” can obscure the transitions between advice, fear, and choice.

This is not a claim that everyone acts freely under all circumstances. Threats, dependency, and coercion can sharply constrain available choices. The inquiry should identify those constraints rather than assume they are absent. But advice, anticipated disapproval, and actual coercion should not be casually merged.

The self-check is therefore symmetrical: am I giving other people more authority in my account than they actually exercised? Am I claiming a shared decision that I made alone? Am I presenting my interpretation as something I directly observed?

6. From empathy to confrontational logic

Within this discussion, our use of snake wisdom concerns discerning inquiry where evasion or deception may be present. “Confrontational logic” means bringing a claim into contact with its particulars, its implications, and potentially contrary evidence. It need not begin with an accusation or an aggressive manner.

Start by understanding the account: “When you say they, who do you mean?” Then establish its content: “What was actually said?” Then examine the connection: “How did that become a decision for both of us?”

If the answers conflict, identify the conflict without pretending to know its cause. “Earlier you said everyone agreed. Now you are describing one person’s advice. Which account should I rely on?”

A useful correction may follow. The speaker might recognize an exaggeration or explain a missing detail. Alternatively, repeated changes in the account may leave a material claim unsupported. We can state that limitation directly without claiming certainty about the speaker’s inner motives.

The inquiry should also remain capable of ending. Once a relevant question has been answered adequately, continually moving the standard can turn clarification into pressure. Asking for specificity does not confer an unlimited right to inspect someone’s messages, identify confidential advisers, or force a defense of every private decision.

Someone can decline an invitation because they do not want to attend. They do not need an external “they” to make that preference permissible. A specific allegation about another person, however, asks the listener to accept something further and can reasonably invite questions about its basis.

7. Unclear, false, misleading, and lying are different

The philosophical literature distinguishes the falsity of a statement from the speaker’s belief and intentions. A traditional account of lying concerns saying something believed false with the intention that another accept it as true; definitions remain contested. Deception is broader than simply uttering a false sentence.[3]

For our purposes, the distinctions prevent premature conclusions:

SituationIllustration in this hypothetical
Unclear reference“They objected,” with no shared understanding of who is meant.
Mistaken reportA speaker sincerely remembers three objections when there were two.
Misleading presentationReporting one objection while omitting others’ support in a way that suggests unanimity.
Deliberate fabricationInventing an objection the speaker knows nobody made, intending the listener to believe it occurred.

A misleading presentation need not be intentional; that is another question to investigate. The practical standard is to say what the evidence supports: unclear, contradicted, incompletely reported, or knowingly fabricated when there is a basis for that stronger conclusion.

8. The same discipline applies beyond relationships

Public language contains the same moves: “they control everything,” “we built it,” “those people want this.” A demographic category may be treated as if it were a committee with a common plan. A few examples may be presented as the character of an entire population.

In racist or white-supremacist reasoning, this inquiry can expose unsupported movement from racial classification to collective intention, personal credit, or collective blame. Ask who is included, what is attributed, and what evidence connects the two.

Precision does not prevent us from evaluating a group’s expressly racist program or people’s knowing support for it. It helps distinguish those supported claims from an assertion about every person who shares a demographic characteristic.

The standard must also govern our own favorable “we.” We cannot demand exact boundaries for an opposing group while letting our preferred collective expand whenever credit is available.

9. A short practice for ordinary conversations

  1. Locate the reference. Who is included, and by what description?
  2. Specify the claim. Is this about words, beliefs, predictions, actions, or authority?
  3. Separate the evidence. What was observed directly, reported by someone else, or inferred?
  4. Check the collective connection. Did these people act together, agree, or authorize a representative?
  5. Calibrate the conclusion. What is supported, contradicted, or still unknown?
  6. Turn the questions inward. Would I accept the same standard for my own “I,” “we,” and “they”?

These questions are prompts, not a demand to recite a checklist during every disagreement. Often one gentle question is enough.

The larger purpose is relational understanding. We want to know whose experience is being described, whose choice is being made, and whose authority is being invoked. Sometimes the answer reveals deception. Sometimes it reveals a misunderstanding. Sometimes it gives someone permission to replace an intimidating “they” with an honest “I.”

Clarify the people. Clarify the claim. Check the connection.

Video transcript

Transcript of the shortened narration used in the accompanying video.

“They told me not to.”

In a relationship, that sentence can put a whole crowd between two people. But who is actually there?

One friend? Several relatives? Or an imagined audience?

“Who do you mean? What did they actually say?”

We begin with empathy: trying to understand.

Then comes ontology: what does the word refer to? A person, a collection, or an organized group?

And evidence: did several people offer advice, or did one person claim to speak for everyone?

An undefined “they” can become a ghost: something we fear or obey without understanding what it names. That does not prove nobody is there. It means we need clarity.

Now try “we.”

“We decided.”

Did both of us agree? Did one person assume? Was anyone authorized to decide for the others?

A pronoun cannot create agreement.

And check “I.”

What did I observe? What am I assuming? Am I presenting my choice as something “they” made me do?

Here, our snake wisdom brings in confrontational logic: examining claims when confusion, evasion, or deception may be involved.

Vagueness alone does not establish a lie. Someone may be uncertain, afraid, or protecting a confidence. Respect privacy while asking what supports the claim.

This applies to public claims about races, nations, and crowds—and to our own relationships.

Clarify the people. Clarify the claim. Check the connection.

Before arguing with “they,” or speaking for “we,” understand who you mean.

Sources and further reading

The sources below provide philosophical background. The relationship examples, “ghost” metaphor, and practical sequence are this article’s applications; these references do not establish a validated method of lie detection.

  1. “Indexicals,” Stanford Encyclopedia of Philosophy. Background on context-sensitive reference and the distinction between indexical and other uses of pronouns.
  2. “Collective Intentionality,” Stanford Encyclopedia of Philosophy. Background on shared intentions and the difference between acting together and an aggregation of individual intentions.
  3. “The Definition of Lying and Deception,” Stanford Encyclopedia of Philosophy. Background on contested definitions of lying, belief, intention, and deception.

Part of Intend to Relate’s Philosophical and Theoretical Foundations. Companion discussions: “Who Is ‘We’?” and “Who Acted?”


Explore Intend to Relate’s concepts and tools at intend2relate.com. Find further writing and explainers at our Ghost publication.