Thought Blocking: Examples, the TLC Rule, and How to Code It

Thought blocking is a train of speech that stops before the thought is finished, and does not come back. Andreasen’s Scale for the Assessment of Thought, Language, and Communication (TLC) calls the item blocking and defines it as the “interruption of a train of speech before a thought or idea has been completed. After a period of silence, which may last from a few seconds to minutes, the person indicates that he cannot recall what he had been saying or meant to say” (Andreasen, 1986).

The definition carries a rule that is easy to miss, and it is the whole key to coding the sign: the rater may not infer blocking from a pause. The speaker has to say the thought is gone.

What is thought blocking?

Thought blocking is the interruption of a train of speech before a thought or idea has been completed, followed, after a silence of seconds to minutes, by the person indicating that they cannot recall what they had been saying or meant to say (Andreasen, 1986). Two elements are required: the interruption, and the report that the thought is lost. A pause that ends with the sentence resuming is a pause; a pause that ends with “I’ve lost it” is blocking.

The TLC then adds the coding rule that distinguishes it from every other item on the scale. Blocking “should only be judged to be present either if a person voluntarily describes losing his thought or if upon questioning by the interviewer, the person indicates that that was his reason for pausing” (Andreasen, 1986). No other TLC item makes the speaker’s own report a condition of rating. Derailment, tangentiality and incoherence are rated from the speech alone; blocking is rated from what the speaker says about the speech. The full set of items is covered in our guide to the Thought, Language and Communication scale.

The reason for the rule is that silences have many causes. People pause to think, to choose a word, to decide whether to say something, because they are distracted, or because they are tired. None of those is blocking, and from the outside they look identical to it. Only the speaker knows whether the thought was lost or withheld.

What do thought blocking examples sound like?

Thought blocking sounds like a sentence that stops partway, a silence, and then the speaker reporting that the thought is gone. The replies below are synthetic.

Interviewer: What happened at your sister’s on Sunday?

Speaker: We had lunch and then she wanted to talk about the, about the… [silence, about eight seconds] …I’ve lost it. It’s gone. I don’t know what I was going to say.

Interviewer: You were telling me about the letter.

Speaker: The letter came on Tuesday and it said that the… [silence, about fifteen seconds]

Interviewer: You stopped. What happened just then?

Speaker: It went. The whole thing. I had it and it went.

Interviewer: How did the appointment go?

Speaker: Fine. She asked about the sleeping and I told her… [silence, about five seconds] …sorry, what was the question?

The first two are blocking under the TLC rule: in the first the speaker volunteers that the thought is lost, in the second they confirm it when asked. The third is a judgment call. The speaker has lost the thread but asks for the question rather than reporting a lost thought; a careful rater asks “what happened just then?” and rates blocking only if the answer is that the thought went. That question is the coding instrument.

A severity rating counts occurrences across the interview: mild for one instance, moderate for two to four, and severe for five or more (Andreasen, 1986). Like the other items in the second half of the TLC, blocking has no extreme anchor. Note that the SANS alogia version of the item, which is the one Tagaroo curates, uses the SANS 0 to 4 scale with an extreme level, so the count anchors here are the TLC’s, not the SANS’s.

What is the difference between thought blocking and poverty of speech?

Thought blocking is an event inside otherwise flowing speech, while poverty of speech is a restriction in the amount of speech throughout the interview; a long pause that ends with the sentence resuming is neither (Andreasen, 1986). The SANS alogia subscale rates blocking alongside poverty of speech, poverty of content and increased latency of response as four separate items, which is the clearest statement that they are different things (Andreasen, 1984).

Sign	What happens	Speaker's report needed?	Where it is rated
Blocking (thought blocking)	Speech stops mid-thought; after a silence the thought is reported lost	Yes, by TLC rule	TLC item 16; SANS alogia
Poverty of speech	Replies are brief and unelaborated throughout	No	TLC item 1; SANS alogia
Increased latency of response	Long delay before replies begin, then a normal reply	No	SANS alogia
Ordinary pause	Silence, then the same sentence resumes	No; not rated	Not rated
Derailment	Speech continues but slides onto another idea	No	TLC item 6

The boundary with increased latency of response is the one coders most often blur. Latency is a delay before a reply starts; blocking is a stop after it has started, inside the thought. The boundary with derailment is whether speech continues: in loose associations the train keeps moving onto other tracks, whereas in blocking it stops. The impoverished-speech items as a group are covered in our guide to the SANS alogia subscale.

How common is thought blocking, and is it specific to schizophrenia?

Thought blocking is rare and not specific to schizophrenia: Andreasen’s TLC says that blocking, like neologisms, is “so infrequent as to be of little use in assessing most patients,” and the same paper notes that behaviors such as blocking “also occur in other psychiatric disorders such as mania or depression, and they also occur in the speech of people who do not meet the criteria for any psychiatric diagnosis, particularly when they are fatigued or stressed” (Andreasen, 1986). The TLC’s introduction states that “some TLC disorders considered to be very important in the past, such as neologisms or blocking, are in fact so infrequent as to be of little use in assessing most patients” (Andreasen, 1986). The TLC’s scoring summary also places blocking in its “less pathological” column, alongside circumstantiality, loss of goal, perseveration, echolalia, stilted speech and self-reference, rather than among the items it treats as more suggestive of severe psychopathology.

That matters for how the sign is taught. Thought blocking is a familiar textbook sign because it is dramatic and easy to describe, and the same is true of neologisms. In practice a coding team will meet both only occasionally, and the frequent items (pressure of speech, derailment, poverty of content) are where agreement is built. When blocking does occur, it was rated reliably in that sample, at a weighted kappa of 0.79, though the sample’s base rate for blocking is not reported in the paper (Andreasen, 1986).

How do you code thought blocking from a transcript?

Code blocking as a span from the point where speech stops to the speaker’s report that the thought is lost, and do not code it at all unless that report is in the transcript. This is instance-level coding, the approach in evidence mode vs instance mode annotation, with one twist: the evidence is the speaker’s own words about the pause, not the pause.

A synthetic exchange with the decisions marked:

Interviewer: Tell me about the new flat.

Speaker: It’s on the third floor, there’s a lift, and the kitchen has… [silence, about ten seconds] …no, it’s gone. I had it. [Blocking: span from “the kitchen has” to “I had it”; the speaker volunteers the lost thought] Anyway, the kitchen. It’s small.

Interviewer: And the neighbours?

Speaker: [silence, about six seconds] Quiet. Mostly quiet. [Not blocking: a delay before the reply begins, then a normal reply; code increased latency of response if rating the SANS, nothing on the TLC]

Three rules keep it honest. Transcripts must preserve pauses with their approximate length, because a transcript that strips silences has removed the sign; clean-verbatim conventions that drop pauses are wrong for this item. Where the speaker did not comment on the pause and the interviewer did not ask, code nothing and flag the span as uncodable rather than inferring. And count instances across the whole interview for the severity band; one clear block is a mild rating, not a feature of the person’s speech (Andreasen, 1986).

With the rule applied, the item that textbooks make mysterious becomes one of the simplest to code: a stop, a silence, and the speaker’s own sentence saying the thought is gone.

What are the most common mistakes when coding thought blocking?

Common coding errors include inferring blocking from a silence, confusing it with latency of response, and rating it from transcripts that have had their pauses removed. Each has a fix:

• Inferring blocking from a pause. The TLC rule requires the speaker to describe or confirm losing the thought (Andreasen, 1986). Ask, or do not code.
• Coding a slow start as blocking. A delay before the reply begins is increased latency of response on the SANS, not blocking (Andreasen, 1984).
• Coding from a cleaned transcript. Clean verbatim removes pauses. For this item, use a transcript that keeps silences and their lengths.
• Treating a word-search as blocking. “The, the, what’s the word” is searching, and the thought is still there. Blocking is the thought going.
• Rating from the most striking instance. Severity is a count: once, two to four, five or more (Andreasen, 1986).
• Reading a diagnosis into one event. The TLC calls blocking too infrequent to help with most patients and places it in the less pathological column (Andreasen, 1986).

Pick the instrument for the question, not the other way round. Thought blocking is the one TLC item that cannot be rated from the speech alone, and the discipline is simple: when the speech stops, ask what happened, and code only what the speaker tells you.

References

• 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 Negative Symptoms (SANS). Iowa City: University of Iowa.
• Andreasen, N. C. (1979). 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

If you code impoverished or disorganized speech from transcripts, Tagaroo curates the SANS alogia subscale, which includes a Blocking item, and the TLC thought, language and communication items as span-level annotation schemes, with inter-rater reliability computed as your coders work. The TLC’s speaker-report rule and count anchors are Andreasen’s (1986) and are not yet encoded in the SANS scheme, so apply them from this guide. The 1979 definitions and reliability study that precede the TLC are Andreasen (1979).