Module 2: Hearsay — Definition, Non-Hearsay & Exemptions

Hearsay is an out-of-court statement, made by a declarant, offered to prove the truth of the matter asserted. All three elements must be present — remove any one, and the statement isn’t hearsay at all.

The rule exists because an out-of-court statement, offered for its truth, can’t be tested the way live trial testimony can. Evidence scholars identify four real, distinct hearsay dangers baked into any assertion: the declarant’s perception might have been flawed (did they actually see what they claim to have seen?), their memory might have faded or distorted the event, their narration might not accurately capture what they actually meant, and their sincerity might be compromised by a lie. At trial, cross-examination, the oath, and the fact-finder’s own observation of the witness’s demeanor all work together to expose each of those four risks in real time. An out-of-court statement, repeated secondhand, strips all of that away — nobody can cross-examine the original declarant about what they actually perceived, remembered, meant, or believed. That’s the real reason the hearsay rule exists, and it’s also exactly why so many exceptions and exemptions exist too: wherever circumstances substitute for one or more of those four safeguards (a statement made under such pressure that fabrication is unlikely, for instance), the law becomes more willing to let the statement in anyway.

Example. A witness testifies that a bystander yelled “the brakes feel broken!” moments before a crash, offered to prove the brakes actually were broken — a classic hearsay problem, since it’s offered specifically to prove the truth of what the bystander asserted.

The same words can be non-hearsay depending on purpose: effect on the listener (showing notice or reaction, not truth), verbal acts (legally operative language like contract terms), and circumstantial evidence of state of mind (proving the declarant’s own mental state, not the statement’s literal truth).

Subramaniam v. Public Prosecutor, [1956] 1 W.L.R. 965 (Privy Council), is the classic case teaching effect-on-the-listener reasoning. A man was charged with unlawful possession of ammunition during Malaya’s communist insurgency, and his defense was duress — he claimed armed terrorists had threatened to kill him unless he carried the ammunition for them. At trial, he wasn’t permitted to testify about the specific threats the terrorists had made to him, on the theory that his account of what they said was inadmissible hearsay. The Privy Council reversed, explaining that the threats weren’t being offered to prove the terrorists actually intended to carry them out — they were offered to show their effect on the defendant’s own state of mind, which was the entire basis of his duress defense, regardless of whether the terrorists were bluffing or serious. The case remains the standard citation for the principle that when the fact of a statement having been said (and its effect on the person who heard it) is what matters, rather than whether the statement was actually true, hearsay simply doesn’t apply at all.

Example. That same “brakes feel broken” warning, offered instead only to show the driver had notice of a problem before continuing to drive — not to prove the brakes were actually broken — becomes non-hearsay, since the evidence’s value no longer depends on the statement being true.

Rule 801(d)(1) treats certain prior statements by a testifying witness as non-hearsay: prior inconsistent statements given under oath at a prior proceeding, prior consistent statements rebutting a fabrication charge, and prior statements of identification.

California v. Green, 399 U.S. 149 (1970), addressed the real constitutional worry lurking behind treating a prior statement as substantive evidence: a witness testified at a preliminary hearing, under oath and subject to cross-examination, that the defendant had supplied him with marijuana; at trial, the same witness recanted and claimed he had been high on LSD at the time and couldn’t recall the truth. The prosecution introduced the earlier preliminary-hearing testimony as substantive evidence, and the defendant argued that doing so violated his Sixth Amendment right to confront the witness against him, since the witness’s in-court account no longer matched it. The Supreme Court held there was no Confrontation Clause problem, because the witness was actually present at trial and subject to full cross-examination there about the earlier statement and any inconsistency — the defendant could ask the witness anything about why his story had changed. The case is the reason Rule 801(d)(1)(A) requires the declarant to actually testify at trial and be subject to cross-examination about the prior statement — that live presence is precisely what makes treating the earlier statement as substantive evidence constitutionally sound.

Example. A witness testified under oath at a deposition that the light was red, but now testifies at trial that it was green — the deposition statement, inconsistent with current testimony and given under oath at a formal proceeding, can be used substantively, not just to attack credibility.

A party’s own out-of-court statement, offered against that same party, is exempted from the hearsay definition entirely — extending to adoptive admissions, statements by a party’s agent/employee (within scope), and co-conspirator statements (during and in furtherance).

Mahlandt v. Wild Canid Survival & Research Center, Inc., 588 F.2d 626 (8th Cir. 1978), pushes the opposing-party-statement exemption further than intuition might suggest. A child was injured after apparently being bitten by a wolf kept by a wildlife organization; the organization’s own employee, arriving on the scene after the fact and without having personally witnessed what happened, left a written note stating “Sorry I did not see your fence, the wolf got loose and injured a child that came into the enclosure” and later told a board meeting the wolf had “bit a child.” The organization argued the note and statement shouldn’t be admissible against it because the employee lacked personal knowledge of the actual event when he made them. The Eighth Circuit held they were nonetheless admissible as opposing party statements, because Rule 801(d)(2) doesn’t require the declarant to have had personal knowledge of the underlying facts at all — only that the statement was made by the party’s own agent, within the scope of that employment, on a matter related to it. The case is the reason personal knowledge, unlike with an ordinary lay witness, simply isn’t a requirement for an employee’s statement to be used against the employer.

Example. Peter tells a mechanic, weeks before a lawsuit, “the brakes had been feeling off for weeks” — his own statement, offered against him, exempted from hearsay entirely. Now compare: a company’s delivery driver tells a customer “our trucks have brake problems,” a statement made within the scope of his job — also treated as an opposing party statement, this time against the company.

When an out-of-court statement contains another embedded out-of-court statement (like a written report quoting what someone said), each layer needs its own independent hearsay exception or exemption before the combined evidence is admissible.

Johnson v. Lutz, 253 N.Y. 124 (1930), remains the classic case teaching exactly why a business record isn’t automatically admissible just because it’s a business record. A police officer prepared an accident report shortly after arriving at the scene, recording what bystanders (not the officer himself) told him about how the collision happened; the report was offered as a business record, since it was regularly kept in the ordinary course of the police department’s business. New York’s highest court held that the business records exception covers the officer’s own firsthand act of recording the report, but it doesn’t automatically extend to the separate, embedded layer of what the bystanders told him — those bystanders had no business duty to report accurately, and the business records exception’s reliability rationale (a duty to keep accurate records as part of one’s job) simply doesn’t reach a stranger’s secondhand account folded into someone else’s official record. The case remains the standard citation for why an official report reciting an outsider’s hearsay account needs its own separate justification for that specific layer, even where the report itself easily qualifies as a business record.

Example. A police report states: “the bystander said the blue car ran the red light.” Two layers exist — the bystander’s original statement, and the officer’s own written repetition of it — both need independent justification, not just one.

Conduct intended as a communicative assertion (like pointing to identify a suspect) is treated like a statement for hearsay purposes. Conduct with no intent to communicate anything (like simply running away) generally is not.

United States v. Zenni, 492 F. Supp. 464 (E.D. Ky. 1980), is the classic case testing the boundary between assertive and non-assertive conduct in verbal form. During a raid on premises suspected of illegal bookmaking, federal agents answered the phone repeatedly, and unknown callers on the other end asked the agents to place bets on various sporting events. The government offered the agents’ testimony about those calls to prove the premises were being used for bookmaking, and the defendant argued the callers’ own words were inadmissible hearsay. The court held the calls weren’t hearsay at all, because the callers weren’t asserting anything — nobody placing a bet is trying to communicate “this is a bookmaking operation”; they’re simply acting on their own belief that it is one, which is exactly the kind of implied, non-assertive behavior that falls outside the hearsay definition even when it takes the form of spoken words. The case remains the standard citation for why the key question isn’t whether words were spoken at all, but whether the speaker actually intended those words as a communicative assertion of the fact they’re now being offered to prove.

Example. A witness at a lineup points directly at one specific person to identify the perpetrator — intended as a communicative assertion, treated like hearsay if offered for its truth. Now compare: a suspect simply flees when police arrive — not intended as any kind of assertion, generally not treated as hearsay even if offered to suggest consciousness of guilt.

Hearsay requires a human declarant. A purely automated record — with no human input or review at the time of recording — generally falls outside the hearsay definition entirely, regardless of the purpose it’s offered for.

United States v. Lizarraga-Tirado, 789 F.3d 1107 (9th Cir. 2015), is a modern illustration of the same principle applied to satellite technology. A border patrol agent used a handheld GPS device to mark the coordinates where he arrested a defendant for illegal reentry, and the government later used Google Earth to place an automatically generated pin, labeled with those GPS coordinates, on a satellite image showing the arrest location was on the U.S. side of the border. The defendant argued the pin’s label was an out-of-court assertion by whoever created it. The Ninth Circuit, in an opinion by Judge Kozinski, held the pin wasn’t hearsay at all, because Google Earth generated it automatically from the GPS coordinates without any human ever reviewing, selecting, or asserting the specific location — there was simply no human declarant making a statement for the hearsay rule to apply to. The case is a real, contemporary reminder that as more evidence comes from automated systems (GPS devices, breathalyzers, computer logs), the threshold question always comes back to the same one Rule 801 has always asked: did an actual person make this assertion, or did a machine simply generate it on its own?

Example. A thermostat automatically logs temperature readings every hour with no human review — offered to prove the actual temperature at a given time, this generally isn’t hearsay at all, since there’s no human declarant making an assertion, though it may still need proper authentication.