Concept
Common Applications
Present sense impression
Common Applications
Rule 803(1) reaches “a statement describing or explaining an event or condition, made while or immediately after the declarant perceived it.”1 Its three elements (a content limit, a perception requirement, and a timing limit) operate on real statements, and real statements of this kind cluster in three settings: emergency calls, accounts that pass through several speakers before they reach the courtroom, and words captured by a recorder or typed into a phone.
Each setting strains a different element. An emergency call usually follows the event by minutes, so the fight centers on timing and on the way an operator’s questions shape what the caller says. A relayed statement raises a problem of structure, because the exception can qualify one layer of a chain while a different rule must qualify every other layer. A recording or electronic message fixes what the speaker said, which removes one doubt, and it leaves open who the speaker was, what the speaker actually saw, and how long the speaker had to compose the words. In each setting the useful question is which element the facts press hardest and which hearsay danger that element guards against. We begin with the emergency call, where all three elements usually line up.
911 Calls
A 911 call fits Rule 803(1) when the caller’s words describe or explain what the caller was watching and were spoken while or immediately after the caller perceived it.2 People who dial for help usually report something they have just seen, so the three elements tend to line up. The caller wants responders sent at once and has every reason to say what is happening and little reason to compose a story, and the gap between event and call is commonly a matter of minutes. The call is also recorded, so no one has to reconstruct the words from memory. Courts accordingly treat calls that report crimes as they unfold, or just afterward, as a standard application of the exception.3
Admitting the recording does not admit the caller’s words, a point students often overlook. The agency’s recording may qualify as its own record, but a citizen who phones in is under no duty to report, so the public-records exception does not reach what the citizen said.4 The caller’s statement is a second layer of hearsay inside the recording and needs an exception of its own, usually the present sense impression or the excited utterance. Calling the evidence “a 911 call” therefore settles nothing, and each call must be tested against the elements, as the decisions below show.
In United States v. Boyce, the caller, Portis, reported a domestic battery to a 911 operator and never testified, so the jury heard only the audio recording and a transcript.5 The district court admitted the call as both a present sense impression and an excited utterance.6 Portis told the operator that Boyce had “just” hit her and that she had “just” run upstairs to a neighbor’s house, and the Seventh Circuit read those words as showing that she called nearly immediately, which is consistent with other circuits’ reading of the exception.7 The caller’s own vocabulary thus fixed the interval from inside the recording, with no call log or outside witness needed.
The defense attacked the content element instead, because Portis mentioned a weapon only after the operator asked about one. The court answered that a speaker can respond to an operator’s questions without calculated narration, and it noted that the operator asked only about a weapon while Portis herself added “a gun.”8 The court conceded that answering questions, together with her run to the neighbor’s house before calling, could raise the chance of calculated narration. It left that point open because the call was admissible as an excited utterance in any event.9 For emergency calls, then, the operator’s questions are one circumstance the court weighs on the narration issue and do not disqualify a call by themselves, and the excited utterance exception stands behind Rule 803(1) when that question is close.
An unidentified caller cleared the perception element on a thin record in Miller v. Crown Amusements, Inc. Police received a 911 call from the caller at a pay phone at a roadside station at 12:13:53 p.m.10 The court thought it a difficult question whether the caller had seen the accident, and it recognized that a judge should hesitate to rest on an unidentified bystander’s statement alone. It still found perception by a preponderance of the evidence from the statement and its circumstances, because the caller “noticed” the truck as it “sideswiped” a car parked on the road.11 The recording left no uncertainty about what the caller said, which heightened the reliability of the evidence.12 On timing, the court invoked the Advisory Committee’s observation that precise contemporaneity is rarely possible and a slight lapse is tolerated, and it found the call substantially contemporaneous because it came, in all likelihood, less than ten minutes after the accident.13 That interval is longer than the “just” in Boyce, but it was spent reaching a telephone.
A call can be prompt, recorded, and vivid and still fail, because the proponent must show that the caller saw the event. Bemis v. Edwards supplies the failing side. Gary Estep’s 911 call described police beating Bemis, and the district court refused to admit it for lack of foundation after concluding from the tape that Estep was relaying what others in his house were seeing and telling him.14 The Ninth Circuit accepted that the statement was nearly contemporaneous and left little room for reflection, yet held that it did not satisfy the further requirement that the declarant have personal knowledge of the events described, and it placed on the proponent the burden of proving that perception by a preponderance of the evidence.15 The court recognized that a caller’s proximity to the scene ordinarily supplies circumstantial evidence of firsthand knowledge, and that a caller who watches a violent arrest could qualify under either exception, but the foundation still had to be shown from the call and its surroundings.16 Set beside Crown, the two calls show the same recording working in opposite directions. In Crown the caller’s words and circumstances pointed to the caller’s own eyes, while in Bemis the tape pointed to a relay, so the feature that removes doubt about what was said can also expose how little the speaker personally knew.
The second step decides most disputes, because a 911 call is rarely attacked on what was said and often attacked on whether the caller saw it. The fifth step matters because the two exceptions measure content and time differently, as the earlier comparison showed.
The Supreme Court borrowed the reasoning behind this exception outside the evidence setting in Navarette v. California, which asked whether an anonymous 911 report that a truck had run the caller off the road was reliable enough for an officer to credit.19 That decision is a Fourth Amendment ruling about police action, and it does not decide whether any statement is admissible. The majority observed that police found the truck about eighteen minutes after the call, some nineteen highway miles south of the reported location, and reasoned that this timeline suggested the caller reported the incident soon after being run off the road.20 It then drew on the Advisory Committee’s rationale for Rule 803(1): a statement about an event made soon after perceiving it is especially trustworthy because the substantial contemporaneity of event and statement makes deliberate or conscious misrepresentation unlikely, and 911 calls that would otherwise be inadmissible hearsay have often been admitted as present sense impressions or excited utterances.21 The majority added that features of the 911 system, including recording and the ability to identify and trace callers, give a false tipster reason to think twice, while cautioning that tips in 911 calls are not per se reliable.22
The dissent doubted that either exception fit the call at all. It described the classic present sense impression as the recounting of an event occurring before the declarant’s eyes as the declarant speaks, and it observed that this caller had time to note a license number, stop the car, copy the number down, and dial the police, which left plenty of time to dissemble or embellish.23 It added that even if less than true immediacy would do, the Advisory Committee’s own Note records hesitancy in accepting an unidentified declarant’s statement alone as sufficient proof of the reported event.24 The dissent also rejected the idea that use of the 911 system vouches for the caller, reasoning that the tipster’s belief in anonymity, and not the actual ability to trace the call, would govern the tipster’s behavior.25
Read together, these opinions show where a 911 call comes under attack. A call that supplies its own timing, like Portis’s “just,” answers the immediacy objection from inside the recording, while a call placed after the caller had time to stop, copy a number, and dial, as the Navarette dissent described, invites that objection. An unidentified caller invites the perception objection, which Crown answered with the caller’s own words and circumstances and which the Navarette dissent pressed through the Advisory Committee’s note of hesitancy. A call to an operator tied to law enforcement can also raise a separate Confrontation Clause question in a criminal case, and some calls come in under the emergency doctrine, a subject for a later section.26
Statements Relayed Through Others
The text of Rule 803(1) speaks of one declarant who perceived one event.28 Proof in real cases often reaches the courtroom through several people: a bystander tells a colleague, the colleague tells a supervisor, the supervisor writes it down, and a witness reads the writing to the jury. Each retelling is an out-of-court statement offered for its truth, and Rule 805 therefore requires every layer of the chain to fall within an exclusion or an exception.29 The present sense impression exception can supply the basis for one layer or for several, an opposing party’s own words can supply the first layer, and a layer the exception does not reach needs a basis of its own. The characteristic error is to assume that qualifying one link qualifies the chain.
The layer-by-layer approach follows from what each speaker adds. Every person who passes a statement along brings a separate risk of mishearing, forgetting, or shading it, and examining the last speaker reveals nothing about the first. The same reasoning fixes what each link must show: a later speaker must have perceived the earlier speaker’s words, and need not have perceived the event the earlier speaker described. If every exception had to reach back to the original declarant, a chain could almost never be admitted, and Rule 805 would have little left to do.30 Each link is therefore measured by what that speaker reports, using the same content, perception, and timing elements that govern a single statement.
United States v. Ferber shows a chain in which an admission carries the first link and the exception carries the next two. The government sought to prove that Ferber had given Merrill Lynch information about an oral examination, and its evidence was a colleague’s handwritten notes of a report from an employee who had spoken with Ferber.31 The notes were hearsay because they were offered to prove that Eisenberg accurately recounted, and Carey accurately wrote down, what Ferber said.32 The statement thus ran from Ferber to Eisenberg to Carey, passing through the heads of two people whose perception and bias could not be tested by cross-examination even though both testified.33
The first link was Ferber’s own words to Eisenberg, which came in as an opposing party’s admission. The court noted that such an admission is still generally treated as an exception to the hearsay prohibition and found a Rule 805 analysis prudent.34 The later links each had to qualify on their own, and the court found that they did. The level of detail in Eisenberg’s answers supported the inference that she reported to Carey with virtual immediacy and had no conceivable reason to falsify, since Merrill Lynch and Ferber were, in the court’s phrase, “as thick as thieves.”35 Carey’s testimony and the notes themselves showed that he wrote as she spoke or immediately afterward, and each of them had actually perceived, by hearing, what the other said.36
Ferber objected that Carey had not perceived the conversation between Ferber and Eisenberg. The court answered that the objection misunderstood the analysis: the notes were a present sense impression of what Eisenberg said to Carey, not of what Ferber said to Eisenberg, and a separate and independent basis supported each link.37 The admission relieved the government of any need to show contemporaneity at the first link, and the exception then qualified each later link by asking only whether that speaker perceived what the prior speaker said.
A telephone call presents the chain in its simplest form, because the listener can report only what the voice on the line said. In Nuttall, a decision from before the Federal Rules took effect, an employee on the telephone with his boss characterized what the boss was saying, both during the call and immediately afterward to his wife, and both parties to the call had died by the time of trial.38 The court admitted the characterizations because they were made substantially at the time he heard the boss’s words, which removed any lapse of memory and made conscious misrepresentation less likely.39 The wife’s lack of knowledge about the call did not matter, since she testified to what her husband said, and he was the one who heard it. Hearing counts as perceiving, so decisions such as United States v. Portsmouth Paving Corp., where a caller told a bystander what the party at the far end had said a few seconds after the call, follow the same pattern.40
The same employee’s later remark shows where the line falls. What Nuttall told O’Hara on arriving at the trainyard was received to show that he felt compelled to come in, but it could not prove that his boss had applied the pressure, and Rule 803(1) would produce the same result.41 The words to his wife came during and right after the call, while the trainyard remark came after the call had ended and the trip had been made, so only the first set was a contemporaneous account of what the boss said.
A chain can also draw on different exceptions at different links. In United States v. Brown, an officer testified that he heard another officer say over the radio that he had been told a man was walking up the street with a gun. Defense counsel conceded that the testimony was admissible because the original declarants’ statements had come in as excited utterances and the second officer’s repetition over the radio was a present sense impression.42 The concession was not a holding, but it shows the structure: one exception supplied the first link, the present sense impression exception supplied the second, and each stood on its own.
The harder case is a chain in which the exception reaches only one stage. In United States v. Medico, a young man who watched bank robbers drive off gave a bank customer the getaway car’s license number, and the customer passed it along moments later to a third person, who wrote it on his checkbook as he heard it. By trial the young man and the customer could not be found, so the third person testified.43 The Second Circuit majority said the testimony met all the specific requirements of Rule 803(1), yet it affirmed on the residual exception because the statement did not satisfy every part of the reasoning behind those requirements.44 Judge Mansfield, dissenting, read the exception more narrowly. In his view it would at best cover the first stage, the young man’s words to the customer, because only the young man was perceiving the getaway he described, and he doubted that a license number and vehicle description describe or explain the “event” at all.45
If the dissent’s reading governs, the stage in which the customer repeats the plate number needs a basis of its own, and without one the chain breaks at that link however well the first link qualifies. Ferber points the other way for that second stage, because the customer perceived by hearing what the young man said and passed it on at once, which is all Ferber demanded of Eisenberg and Carey. Whichever reading a court adopts, each arrow in the chain needs a route, and a different rule, such as the residual exception the Medico majority used, can supply one when Rule 803(1) does not.
Rock v. Huffco Gas & Oil Co. shows a chain failing at its first link. The plaintiffs offered accident reports that Christian and Gardner prepared from what Rock told them about an alleged accident, and the Fifth Circuit called the reports classic hearsay within hearsay, so that both levels needed an exception.46 Rock’s own statements could not qualify, because one report was filed two days after the alleged accident, Rock had time to manipulate the truth, and there was evidence that he had staged the event for personal gain.47 Recasting the reports as recorded recollection did not help either, since that exception reaches only a writer who once had firsthand knowledge of the facts the writing records. What separates Rock from Ferber is the first link. Both involved a written record of a spoken account, but Ferber’s first speaker was an opposing party and the later speakers reported at once to colleagues with no motive to falsify, while Rock’s first speaker was the person with the motive to fabricate and spoke after a delay that gave him time to do it.
The common store incident report shows the same two-step structure in everyday form. A store manager investigating a shopper’s injury writes a report quoting a bystander, who spoke about ten minutes after the accident. Offered to prove how the shopper was acting, the report has two layers: the bystander’s remark, which might be a present sense impression although some courts would find ten minutes too long, and the manager’s written report, which is likely a business record if the store prepares such reports as a regular practice.48 Rejecting either layer excludes the entire report.49 A polished business record cannot cure a first layer that fails the timing element, and a well-timed remark cannot cure a record that fails its own exception.
The third step does the work that separates a chain from a single statement, because it changes the object of perception at every link after the first. The fourth step explains why a failed link is not always fatal, since another rule may cover it, while a link with no basis at all ends the analysis.
Recorded and Electronic Statements
Rule 803(1) says nothing about the channel a statement travels through, so a statement describing or explaining an event or condition, made while or immediately after the declarant perceived it, qualifies whether it is spoken to a friend, dictated into a recorder, or typed into a phone.51 The medium leaves the test untouched and changes only how the elements are proved. A recording or message is a container for the declarant’s words, and those words remain an out-of-court statement that needs an exception of its own whenever they are offered to prove what they say. Two things do shift with the medium. A recording fixes exactly what was said, which removes a doubt that a listener’s memory would leave open. An electronic message can be sent within seconds of the thing it describes, which suits the exception, but it can also be drafted, edited, and posted at leisure, which does not.
A recording made while the declarant watches the event is about as clean a present sense impression as the exception produces. In United States v. Andrews an agent looked through the window of a home, saw food stamps exchanged for cash, and recorded on tape his impressions of what he saw as he was looking. When the defense offered testimony and photographs suggesting that he could not have seen through that window, the government offered the tape.52 The Eleventh Circuit rejected the objections that the tape was self-serving hearsay recorded outside the defendants’ presence, holding that it fell clearly within the exception because the substantial contemporaneity of event and statement makes deliberate or conscious misrepresentation unlikely. The court also found no showing that the agent had a motive to fabricate at the moment he looked in the window. The mechanism is the one that supports the exception generally: words captured while the eyes are on the scene involve no memory to fail and almost no time to plan, and the tape preserves them exactly. The same recorder switched on after the agent had left the street would preserve the words as faithfully and prove nothing about immediacy, because the timing element looks to when the agent perceived what the tape describes.
A recording also changes how much a second witness adds. Under the Federal Rules a present sense impression need not be corroborated, but the hearsay gains trustworthiness and probative force where corroborating circumstances or witnesses are available, and the need for corroborating witnesses is somewhat reduced where a tape recording leaves no uncertainty about the content of the declarant’s statement.53 The reason is that a corroborating witness answers two doubts, what the declarant said and whether the declarant had it right, and a tape answers only the first. With the wording fixed, the dispute narrows to perception and accuracy, and for that dispute a second observer remains useful though never required. The recording can also cut the other way. In the ordinary setting the witness who reports a statement has perceived the event or at least seen circumstances strongly suggesting it, which adds assurance as a by-product, though a general justification for admission is not a requirement.54 When the sponsoring witness only plays a tape or identifies a log, that by-product may be missing, and the judge is left with the recording and its surroundings. Those can be enough, because the court decides under Rule 104(a), without being bound by the rules of evidence, whether Rule 803(1) applies and may treat the statement itself as some proof that the declarant perceived the event, as in Miller v. Crown Amusements, Inc., where a 911 caller’s own words and circumstances showed that the caller observed the accident.55
Electronic messages bring the exception into more everyday use. Witnesses now tweet, text, post, or call 911 to describe crimes as they unfold or just afterward, as in United States v. Boyce, where a 911 call reporting that the defendant had just hit the caller was admitted under Rule 803(1).56 The spread of cell phones and text messaging multiplies the occasions on which people narrate contemporaneous observations to someone else, so the exception may see more frequent application.57 It also affects what counts as immediate, because “immediately after” may fairly be measured by roughly the time it takes to locate a phone and send a text or make a post.58 Speed of the medium still does not decide admission. A witness who sees a car run a red light and hit someone and tweets at once, “Saw a car run a red light and hit someone,” has made a present sense impression, whereas the same words posted an hour later come too late.59 The hour matters because it is long enough for reflection, and the question the timing element asks is whether enough time passed to permit reflective thought.60
Many posts never need the exception at all, because they are offered for something other than their truth. In People v. Valdez a defendant challenged the admission of content from his Myspace page as hearsay, and the appellate court rejected the challenge because the trial court had not admitted the material for the truth of any assertion on the page.61 The same ladder of questions applies to every electronic statement: whether it is a statement offered for its truth, whether it is excluded from the hearsay definition (as an opposing party’s statement, for example), and only then whether an exception such as the present sense impression applies.62 At that last step the electronic setting adds a twist, because the features that make a message quick also make it editable. A declarant can draft a sentence, revise it, and post it within a minute or two, so a short interval on the clock no longer guarantees that the words were unreflective. A platform’s timestamp, where one exists, records when the message was posted and so gives the judge a measurable end point, but it does not record when the declarant perceived the event described. The other end of the interval has to come from the content of the message, from related messages, or from other proof, and the content limit adds a second check, since a post that offers causes or blame reaches beyond what the declarant was perceiving.63
The percipient-witness proposal introduced with the corroboration discussion comes to a head with electronic statements. One proposal would allow a present sense impression in electronic form to be presented only through a witness who was present when the statement was made and who can clarify, vouch for, or even discredit it.64 The worry behind it is that a post typed alone and read later by strangers lacks the listener beside the speaker who, in the ordinary case, can check the account. The answering view holds that existing doctrine already deals adequately with electronic hearsay and that a percipient-witness requirement would undermine the exception’s usefulness, particularly in domestic violence cases.65 That response draws on the fact that the exception already carries limits that keep it from admitting social media hearsay wholesale, and they can be summarized in three points:66
- The event or condition described by the statement must have occurred or existed.
- The declarant must have personal knowledge of that event or condition.
- The declarant must make the assertion while perceiving the event or condition or immediately afterward.
These three points restate the perception and timing demands taught earlier, together with the premise that there was something to perceive. The judge decides whether they are met and may consider the statement itself as some proof, so a message that reveals nothing about its writer’s vantage point or timing cannot lift itself into evidence.67 Rule 803(1) as written imposes no percipient-witness requirement, which leaves these limits and the judge’s findings as the whole of the screening.
The tenant’s text shows why the three limits do the screening work that a percipient-witness rule would otherwise perform. The sister’s testimony authenticates the message but adds nothing about the window, so under the proposal the text would fail, while under the federal rule the judge asks whether the message, its timing, and the recording make perception more likely than not. The forum post fails for reasons no witness could repair, since its words describe earlier events and came after hours of opportunity to reflect. Qualifying under Rule 803(1) still leaves other hurdles, including authentication of the message and, for calls to operators tied to law enforcement in a criminal case, a Confrontation Clause problem, both of which come later under the limits beyond the exception.68