Herat · Afghanistan · Oral recording

A voice from Herat

A preliminary study of a machine-generated transcript of a Persian-language Mushkil Gusha telling attributed to Humaira and preserved in the Margaret Mills collection.

Working status

English archive version, 13 September 2026, based on the initial assessment of 4 September 2026 and a separately supplied short second recording and transcription covering 16:35–17:15. This is neither a citation-ready transcript nor a complete translation of the recording.

Source and verification

What has been examined

Recording title: داستان عامیانه: نذر مشکل گشا / مارگریت ملز؛ گوینده حمیرا
Context: Attributed to Margaret Mills, Humaira and Herat on the basis of the catalogue context and recording title.
Date: The recording date remains uncertain. The date 4 September 2026 refers to transcription, not recording.
Transcript: 651 subtitle blocks covering 00:00:00–00:25:01.860. All blocks were read, but the timestamps were not checked against the audio and the SRT has no reliable speaker attribution.
Method: Initial assessment of an automatic speech-recognition transcript, supplemented by a second short transcription of 16:35–17:15.

The automatic transcript is intelligible in substantial passages but contains phonetic misspellings, damaged Arabic prayer formulas and conspicuous multilingual recognition artefacts. The passages at 04:50–05:18 and 10:29–10:58 are especially corrupt. The initial assessment did not include a complete listening check. Gaps must not be filled from other known versions of the story. Even apparently clear passages remain provisional until checked by a qualified listener against the audio.

Narrative

Provisional outline

Time	Content discernible from the transcript	Limitation
00:09–01:02	An old thorn-gatherer lives with his daughter in a small mud dwelling. He sells the thorns he collects and uses the proceeds to obtain modest food.	خارکن is clearly repeated. There is no basis for translating it as “woodcutter”. No wife appears in the readable opening, but this does not prove that he is a widower.
01:04–03:34	The daughter longs for meat. She smells a dish at a neighbour’s house, goes there repeatedly but receives nothing, and asks her father to bring some home. His further search for thorns is unsuccessful; father and daughter go hungry.	The dish is recognised as something like دلبن or دلمند, perhaps a dialect word or recognition error. It cannot yet be identified securely.
03:39–04:49	A horseman meets the father outside the house. He hears about the man’s distress, asks him to give his hand, recite seven salawāt and close his eyes. The man finds himself in another, desolate landscape and is told to collect pebbles, filling his pockets and clothing.	Individual words are faulty, but the sequence is clear. A visible horseman appears; there is no disembodied voice at this point.
04:50–05:18	The collecting of stones continues.	Heavily damaged passage with English, Korean, Cyrillic and other intrusions; not translatable without checking the audio.
05:19–06:47	Further containers are filled. After again giving his hand, reciting salawāt and closing his eyes, the father is back at home. He pours the stones into a corner of the courtyard. At night the daughter notices light; neither initially understands their value.	The exact meaning of some night-time actions remains open. No fire or neighbour episode is imported from other texts.
06:48–08:47	The father tries to exchange the stones for bread. The baker directs him to a goldsmith. The goldsmith increases his offer while the old man still does not understand the stones’ value. After the transaction each seems to fear that the other may reverse it.	The monetary amounts are uncertain and are not retained as reliable data.
08:48–09:28	Father and daughter become prosperous. The daughter owns jewellery, meets the king’s daughter at the bathhouse and befriends her. She apparently gives her a necklace.	The gift is plausible in the transcript, but the wording requires verification.
09:29–10:22	The father leaves on pilgrimage to Mecca. It is then explained that the helper had instructed him to make a regular offering in the name of Mushkil Gusha. He leaves chickpeas and raisins and tells his daughter to make the offering monthly. Absorbed in her friendship with the king’s daughter, she forgets.	هر برجی appears to mean “every month” here. References to three coins or three bowls are plausible but more seriously damaged.
10:23–11:16	The loss of jewellery and imprisonment form the next section. The father is in prison and is reminded of the Mushkil Gusha offering by the helper. He says that he left provisions at home and does not know whether his daughter performed it.	The transition at 10:29–10:58 is badly damaged. The accusation, arrest and return from pilgrimage cannot be reconstructed reliably.
11:17–12:54	Following the helper’s instruction, the father apparently finds coins in his beard and asks a passer-by to obtain the offering. The man is hurrying to a doctor for his dying son and initially refuses. The son dies. On his way to buy a shroud, he meets the prisoner again, now fulfils the request, and later finds his son alive.	The readable sequence suggests the same man first refuses and later helps. The delivery of the offering and the release remain damaged. No second refusing rider and accident are evident in the readable passage.
12:55–13:23	An intervening question returns to the necklace. A bird is mentioned and the jewellery is apparently recovered; the thorn-gatherer is freed or is again at home.	The exact means of discovery cannot be translated reliably. “Bird” is plausible; crow, nest, reflection or sneeze are not supplied.
13:24–14:17	The revived son reports that a veiled or masked horseman granted him twenty further years of life because his father had rendered Mushkil Gusha a service. This father too is to make the offering in future. The story ends in prosperity and with an association with ʿAli.	Twenty additional years are clearly readable. The precise identification around 14:12 is corrupt but is supported by the following prayer context.

Living practice

Offering, telling and distribution

14:23–15:09 · The offering

The teller describes her own performance of the nazr. Its scale apparently depends on available means: those who have much give more, and those who have little give less. Chickpeas and raisins are named. The story is told while the food is being prepared, selected or cleaned; afterwards the offering is divided onto plates and distributed. A question mentions neighbours or passers-by, but the exact answer is poorly recognised.

A statement at 14:48–14:53 may concern telling for oneself. It is an important point for further checking, but does not yet establish that the story may generally be told without a listener. “I tell it myself” and “I tell it to myself” must be distinguished by listening to the recording. If confirmed, this would also need to be distinguished from Shah’s motif in which the protagonist tells himself his own life story.

15:09–16:35 · The prayer

Nād-e ʿAlī is named explicitly and is followed by a longer Arabic recitation. Repeated invocations of ʿAli are recognisable. Many Arabic formulas are mistranscribed and must not be silently replaced with a standard form and then quoted as the wording of this recording.

16:44–17:03 · Where the prayer and story came from

The separately transcribed follow-up excerpt helps clarify this part of the conversation. On the most probable reading, the dialogue distinguishes between the prayer and the story:

A book source is mentioned for the Arabic prayer.
When asked about the story, the teller appears to say that someone told it to her and that she did not read it from a book herself.
At the same time, it appears to be said that the story also exists in a book. The book is not identified.

از کتاب نخوندم یک نفر بست.
The first part means approximately: “I did not read [it] from a book.” The final words are faulty in the automatic transcript.

This provides possible first-person evidence of one step of oral transmission. It does not establish the age of the tradition or its independence from printed versions; the earlier teller may have known a printed text.

17:15–25:02 · Other vows and stories

The conversation turns to other food vows, including sholeh-zard and a dish recognised as دکچه. Questions follow about occasions, ingredients, repetition and Christian comparison practices. Dates and ages mentioned around 17:27 and 20:20 cannot date the recording until wording and speaker attribution are verified.

Around 21:06–21:28, the speaker appears to say that Mushkil Gusha offerings are performed by both Shiʿa and Sunni Muslims, but especially by Shiʿa. This is the participant’s assessment, not a statistical statement about Afghanistan. Names including Bibi Nur and Bibi Seshanbe occur from about 21:33. The subsequent shift in conversation does not justify treating all named vows as identical. A different story beginning around 22:50 does not belong to the Mushkil Gusha narrative.

Comparative significance

What the recording presently shows

The existing comparison with Lorimer (1919) and Shah (1968) is used here as working context. Their original texts were not completely recollated for this preliminary note.

The thorn-gatherer is explicit. The protagonist is repeatedly called خارکن. This does not establish the wording of every Afghan variant and is not by itself a secure marker of geographic origin.
The recording contains an extensive Mushkil Gusha complex. The daughter’s desire, hunger, miraculous stones, prosperity, the king’s daughter, lost jewellery, imprisonment, remembrance of an offering and rescue are present or discernible despite damaged passages.
Several features differ from Idries Shah’s published form. They include a visible horseman, salawāt, a monthly offering, pilgrimage, the daughter’s forgetting, chickpeas and raisins, and the death and revival of a son.
The invisible stairway is not positively attested. In the readable passage the marvellous journey occurs through giving a hand, prayer and closing the eyes. The damaged section prevents an absolute statement of absence.
Thursday and dates are not securely attested. The positive evidence is a monthly instruction and chickpeas with raisins. With transcript quality this poor, an unsuccessful word search cannot prove absence from the recording.
An Afghan oral transmission path has become a concrete research question. The passage suggests that Humaira learned the story from another person, but establishes no direct source relationship to Idries Shah. Without reliable dating and a chain of transmission, the recording can be treated neither as his source nor as evidence against possible influence from Lorimer.
Shared motifs are not sufficient evidence of descent. A source relationship would require distinctive wording, sequences of motifs or documents demonstrating transmission.

Work remaining

Priority checks against the audio

16:44–17:04: exact wording about the origins of prayer and story, the oral intermediary and the reference to a book.
14:23–15:09: telling alone or before listeners; the relationship between telling, preparing the offering and distributing it.
10:23–11:17 and 12:55–13:24: necklace loss, accusation, imprisonment, the bird and the recovery.
04:40–05:35: the full sequence of collecting and the marvellous journey.
09:29–10:22: monthly rhythm, quantities, pilgrimage and the daughter’s responsibility.
21:02–21:29: the statement about Shiʿa and Sunni practice.

Further machine recognition of short passages may assist comparison, but cannot replace review by a qualified Persian-speaking listener.

The automatic transcript indicates an extensive Persian-language Mushkil Gusha narrative involving a thorn-gatherer and an Islamic votive practice. The subsequent conversation appears to distinguish between a story learned orally and an Arabic prayer obtained from a book. Damaged speech recognition, uncertain dating and an incomplete chain of transmission do not yet permit a conclusion as to whether this version was transmitted independently of earlier printed texts.