Tangentiality vs Circumstantiality: Examples and Coding

Tangentiality vs circumstantiality comes down to one question: does the speaker ever get to the point? In circumstantiality, the reply wanders through tedious detail and eventually arrives at the goal idea. In tangentiality, the reply to a question is oblique or irrelevant (Andreasen, 1986), and in practice the question is not answered. A circumstantial thought process still arrives at the point; a tangential thought process does not. Everything else about the distinction, from the definitions to the reliability figures to the coding rules, follows from that endpoint test.

The two signs get confused because they sound alike in the first ten seconds. The difference only becomes visible at the end of the reply, which is why the interviewer’s patience, and where a transcript is cut, decide what gets coded.

What is the difference between tangentiality and circumstantiality?

The difference between tangentiality and circumstantiality is whether the goal idea is reached. Andreasen’s Scale for the Assessment of Thought, Language, and Communication (TLC) defines circumstantiality as a very indirect and delayed way of reaching the goal idea, weighed down with tedious detail, where the point is eventually made, and tangentiality as replying to a question in an oblique, tangential, or even irrelevant manner (Andreasen, 1986); in practice, the question is not answered. The amount of detail is a feature of circumstantiality, not the test. A short reply can be tangential; a very long one can be perfectly circumstantial.

Andreasen’s project in the late 1970s was to retire the catch-all term “formal thought disorder” and replace it with operational definitions that two raters could apply to the same interview and agree on (Andreasen, 1979a). Tangentiality and circumstantiality are two of the 18 scored items in that vocabulary (the TLC adds two paraphasia items that sit outside the global rating), and the endpoint test is what keeps them apart. The full glossary is covered in our guide to the Thought, Language and Communication scale.

What do tangential speech examples look like?

Tangential speech starts somewhere near the question and moves away from it, so that by the end of the reply the question is still open. The replies below are synthetic, written to illustrate the sign.

Interviewer: How have you been sleeping?

Speaker: Sleep is a strange word. My grandmother had a sleeper sofa, green, from a catalogue. Catalogues were a big thing back then, you’d wait weeks for the van.

Interviewer: What brought you to the clinic today?

Speaker: The 42 bus. It’s been rerouted since the roadworks, past the old cinema, which they’re turning into flats. Everything’s flats now.

Interviewer: Are you taking the medication?

Speaker: Medication comes in those blister packs, and the foil is the same foil as on yogurt lids, which is a thing about packaging I’ve noticed.

In each case the first few words connect to the question (sleep, the journey, medication) and then the reply pivots to something adjacent and keeps going, never returning. That initial connection is what separates tangentiality from incoherence, where the words themselves fail to connect, and from a flat refusal to answer. The speaker is responding. They are just not answering.

A severity rating follows the TLC anchors, which count instances across an interview of roughly 50 minutes: mild when tangentiality occurs once, moderate at two to four times, severe at five to ten, and extreme above ten (Andreasen, 1986). A single tangential reply in a whole interview is a mild rating, not a feature of the person’s speech.

What does circumstantial speech sound like?

Circumstantial speech is a long way round to an answer that does eventually come. The same three questions, answered circumstantially:

Interviewer: How have you been sleeping?

Speaker: Well, we moved the bed last month because the radiator was leaking, and the plumber could only come on a Thursday, which is my late day, so for a week I was on the sofa, and the cat kept me up, but since the bed went back, honestly, about six hours, broken.

Interviewer: What brought you to the clinic today?

Speaker: My sister. Not literally, she’s in Leeds, but she rang on Sunday, she always rings on Sunday, and she said I sounded flat, and she’s usually right about that, so I booked it. The flatness, I suppose, is why I’m here.

Interviewer: Are you taking the medication?

Speaker: The pharmacy changed brands, the new box is blue, I nearly didn’t recognise it, and the pills are oval now, which I don’t like, but yes, every morning with the coffee.

Each reply is over-inclusive. Detail that a direct speaker would prune (the plumber’s Thursday, the colour of the box) is kept in. But every reply lands: six hours, broken; the flatness; yes, every morning. A rater who stopped reading halfway through would code all three as tangential, which is exactly the trap.

Circumstantial speech is also common outside any clinical population. Anxiety, a detail-oriented style, and the wish to be thorough with a doctor all produce it, which is why it is a poor marker of any one condition, even though it leaned toward mania in Andreasen’s sample (Andreasen, 1979b). It describes a route, not a disorder. The TLC’s own mild anchor is a speaker who gives an occasional circumstantial reply but “can get to the point quickly if interrupted and urged to do so” (Andreasen, 1986).

How do you tell tangentiality from derailment and loss of goal?

Tangentiality is an oblique reply to a question; derailment is slippage within the speaker’s own spontaneous speech; loss of goal is a chain of thought that wanders off and never reaches its conclusion; circumstantiality reaches the goal late (Andreasen, 1986). The four drift signs are separated by what triggers the drift and where the speech ends up, and the table lines them up, with derailment vs tangentiality as the pair most often confused.

Sign	Trigger	Does the point get reached?	How it looks in a transcript
Circumstantiality	Any reply or narrative	Yes, late	Long reply, excessive detail, answer at the end
Tangentiality	A question	No	Reply starts near the question, moves away, never returns
Derailment	The speaker's own spontaneous speech	Usually no	Ideas slide onto obliquely related topics mid-stream, without a question prompting it
Loss of goal	A chain of thought	No	Speech runs on without ever reaching the natural conclusion of its own argument

The question is the hinge. The TLC keeps the two apart by definition: “Tangentiality has been defined herein as a different phenomenon in that it occurs as the immediate response to a question” (Andreasen, 1986), whereas derailment happens inside the speaker’s own spontaneous speech. In practice the two co-occur, and a long derailed monologue often begins as a tangential reply. Our coding rule: code the first move by its trigger and the later slippage as derailment.

Flight of ideas, the manic pattern of rapid topic shifts, is not a separate TLC item. In Andreasen’s words, “Flight of ideas is a derailment which occurs rapidly in the context of pressured speech” (Andreasen, 1986). The positive-symptom version of these items, with its 0 to 5 scoring, lives in the SAPS positive formal thought disorder subscale; the two instruments are compared in TLC vs SAPS.

Which conditions show tangentiality and circumstantiality?

Neither sign points to one condition. In Andreasen’s comparison of 113 patients with mania, depression or schizophrenia (the same sample that underlies the TLC reliability figures; Andreasen, 1979b, 1986), circumstantiality was more strongly associated with mania, along with pressure of speech, distractible speech and clanging, while derailment, tangentiality, illogicality, incoherence and loss of goal occurred at similar rates in mania and schizophrenia (Andreasen, 1979b). What did discriminate schizophrenia was the negative side: poverty of speech and poverty of content.

That finding retired an old assumption. Loose associations had been treated as close to pathognomonic of schizophrenia; Andreasen showed they were as common in mania (Andreasen, 1979b). The lesson for anyone coding speech is to describe the sign and leave the diagnosis to the clinician with the whole picture. A transcript full of circumstantial replies is a transcript full of circumstantial replies. The impoverished-speech counterpart, which does lean toward schizophrenia, is covered in our guide to the SANS alogia subscale.

Why is tangentiality harder to rate reliably?

Tangentiality is harder to rate because it asks the rater to judge an absence. Circumstantiality reached a weighted kappa of 0.74 in Andreasen’s reliability study; tangentiality reached 0.58 (Andreasen, 1986). The circumstantial rater has a positive event to point at, the moment the answer arrives. The tangential rater has to be confident the answer would not have arrived had the reply gone on, and that depends on how long the interviewer let it go on.

Two interview habits inflate tangentiality ratings. The first is interrupting: a circumstantial reply cut off at 20 seconds can look like a tangential one. The second is asking closed questions, which produce one-word answers and leave no room for either sign to appear. The TLC bases its ratings on a whole interview of roughly 50 minutes, and its circumstantiality anchors explicitly describe what happens when the interviewer interrupts and urges the speaker to the point (Andreasen, 1986). For recorded interviews, the same applies to transcription: a transcript that truncates long replies will systematically over-code tangentiality.

How do you annotate tangentiality and circumstantiality in a transcript?

To annotate these signs, mark the span from the interviewer’s question to the end of the speaker’s reply, then label it by the endpoint test: circumstantiality if the goal idea appears anywhere in the span, tangentiality if it does not. This is instance-level coding, the approach described in evidence mode vs instance mode annotation, and it makes every rating checkable by a second coder.

A synthetic coded exchange, with the decision at each step:

Interviewer: What did the doctor say at your last appointment?

Speaker: The last one was the Tuesday, no, the Wednesday, because the Tuesday was the dentist, and the dentist is on the same road, which is handy, and she said the bloods were fine but to keep an eye on the iron. [Circumstantiality, mild: the goal idea arrives (“bloods were fine, keep an eye on the iron”) after unnecessary detail]

Interviewer: And how did you feel about that?

Speaker: Feelings are like weather, my dad used to say that, he had a barometer in the hall, brass, you tapped it, and the needle would jump. [Tangentiality, moderate: starts at “feelings”, moves to the barometer, the question is never answered]

Three rules keep the coding honest. Decide the label only at the end of the reply, never mid-way. Record the severity from frequency across the interview, not from the single worst instance, as the TLC anchors require (Andreasen, 1986). And when the interviewer interrupts, note what happened next: in the TLC, a speaker who reaches the point quickly once urged is mildly circumstantial, and one who keeps going after being interrupted rates higher (Andreasen, 1986). If the reply is cut before either outcome is visible, flag the span as uncodable.

With a scheme like this the two signs become something two coders can agree on, and disagreements become informative: a split between circumstantial and tangential on the same span almost always means one coder stopped reading early or the transcript was cut.

What are the most common mistakes in coding tangentiality vs circumstantiality?

The most common mistakes are coding by the amount of detail rather than the endpoint, coding tangentiality from a reply the interviewer cut short, and conflating tangentiality with derailment. Each has a fix:

• Coding by length. A long reply is not tangential because it is long. Read to the end; if the goal idea appears, it is circumstantial (Andreasen, 1986).
• Coding an interrupted reply as tangential. If the interviewer cut in, look at what the speaker did next; if nothing is visible, flag the span rather than guess.
• Merging tangentiality and derailment. Tangentiality needs a question. Drift inside spontaneous speech is derailment (Andreasen, 1986).
• Calling it incoherence. Incoherence is speech you cannot follow at the sentence level. A tangential reply is perfectly followable; it just does not answer.
• Reading a condition into one sign. Circumstantiality leaned toward mania and tangentiality did not separate mania from schizophrenia in Andreasen’s sample (Andreasen, 1979b). Describe the speech and stop there.
• Rating from the worst moment. TLC severity is about frequency across the interview, not the single most striking reply (Andreasen, 1986).

That’s the difference between a number and an interpretation. Tangentiality vs circumstantiality is decided at the end of the reply, by whether the point was ever reached; everything before that is detail the rater has to sit through, which is, after all, the sign.

References

• Andreasen, N. C. (1979a). Thought, language, and communication disorders. I. Clinical assessment, definition of terms, and evaluation of their reliability. Archives of General Psychiatry, 36(12), 1315–1321. doi:10.1001/archpsyc.1979.01780120045006
• Andreasen, N. C. (1979b). Thought, language, and communication disorders. II. Diagnostic significance. Archives of General Psychiatry, 36(12), 1325–1330. doi:10.1001/archpsyc.1979.01780120055007
• Andreasen, N. C. (1986). Scale for the Assessment of Thought, Language, and Communication (TLC). Schizophrenia Bulletin, 12(3), 473–482. doi:10.1093/schbul/12.3.473
• Andreasen, N. C. (1984). The Scale for the Assessment of Positive Symptoms (SAPS). Iowa City: University of Iowa.

If you code disorganized speech from transcripts, Tagaroo curates the TLC thought, language and communication items and the SAPS positive thought disorder subscale as span-level annotation schemes, with the endpoint rule above built into the coding guidance and inter-rater reliability computed as your coders work.