بسم الله وبحمده والصلاة والسلام على خير خلقه سيدنا محمد وآله
Imam Muhammad ibn Ismail al-Bukhari (may Allah be pleased with him) strove to restrict his famous collection to the most authentic hadith traditions, and though we know it as ‘Sahih al-Bukhari,’ its title is actually much longer than that, and begins with ‘al-Jami’ al-Sahih al-Musnad…’
The word musnad here refers to traditions that connect back to the Messenger of Allah (may Allah exalt him and send him greetings of peace) through chains that appear to be fully connected, meaning that every narrator heard the tradition from the person above him in the chain. For example, when every person in the chain uses terms like “so-and-so narrated to us,” or even “[we narrate] on the authority of…”, the chain appears to be connected, and so it qualifies as musnad. Thus the term musnad refers to the outward appearance of the hadith, not its actual reality, because many hadith transmitters use expressions like, ‘[we narrate] on the authority of,’ from people that they did not hear the hadith from. Rather they may have heard this hadith from intermediaries that they dropped out of the chain, and this is called ‘tadlees‘ and is of course blameworthy. Therefore, the term musnad refers to the fact that the hadith appears to be fully connected from the wording of the chain. As for the word sahih, it means rigorously authenticated, therefore Imam Bukhari’s work is a collection of musnad traditions which he rigorously authenticated, and found to be truly connected in reality, as they appear to be outwardly. Of course to qualify as sahih there are other criteria such as the truthfulness and accurate memory of all the narrators in the chain of the hadith.
These authentic musnad hadiths, therefore, are what the Sahih al-Bukhari are all about. They are the asl, the main foundation of the work, the intended content. However, they are not the only content of the work. Imam Bukhari also included in his book 1341 hadiths that are mu’allaq, literally: left hanging. A hadith that is mu’allaq is defined as a hadith in which the author of the work omitted one or more persons from the beginning of the chain (the author’s end of the chain). Examples are: to omit the entire chain (and simply write, ‘The Messenger of Allah said…’), or everyone except the Companion narrating from the Prophet, or everyone except the Companion and the Successor. It could also be the statement of a Companion, not a prophetic tradition, also quoted without any chain. Or it could be to include everyone in the chain except the author’s own shaykh for example. Imam Bukhari included 1,341 mu’allaq hadiths in this collection, however in reality the majority of them are simply found elsewhere in the book with complete chains and he therefore did not bother repeat the chains again. This leaves 160 mu’allaq hadiths that do not have a full supporting chain anywhere else in the book by the count of the great Ibn Hajar al-Asqalani. Ibn Hajar, who would be honored like al-Bukhari before him with title amir al-mu’mineen (Commander of the Faithful) in the field of hadith, collected these 160 mu’allaq traditions in a book which he called Taghliq al-Ta’liq, and strove to find full chains of transmission for these traditions from other sources outside of al-Bukhari’s book. Ibn Hajar later composed the greatest and most celebrated commentary on Sahih al-Bukhari, which he called Fath al-Bari. These mu’allaq traditions are usually found in the beginnings of Bukhari’s different chapters and sections, and are supplementary material, not the intended bulk of the work.
There is another type of hadith classified as mu’allaq however, like the ‘Hadith of Instruments’ with which we are concerned, and this type is very interesting because it contains a fully connected chain of transmission from the author’s own shaykh back to the Messenger of Allah (may Allah exalt him), however, the author’s wording does not give us an indication of audition from his shaykh. For example, the author might say that one of his own teachers “said” (qala) and then mention the teacher’s full musnad chain. This leaves us in doubt as to whether or not the author heard this hadith from the teacher. It is possible that he did not, and that he has omitted the intermediaries, and it is possible that he did hear it from him, but for one reason or another chose not to indicate this. The question here is why al-Bukhari or other authors would do that.
Imam Ibn Hajar studied the mu’allaq hadiths in Sahih al-Bukhari very carefully, and studied other works of Imam Bukhari like al-Tarikh al-Kabir. In his introduction to Taghliq al-Ta’liq he mentioned five reasons in general for al-Bukhari to leave a hadith ‘hanging’:
1) Because it is repeated (it has been quoted elsewhere with its full chain)
2) Because he has already quoted another hadith with full isnad that conveyed the same meaning and then added this quote without a chain for the sake of brevity
3) For clarifying something in the chain about a particular narrator’s taking from another (a technical matter that is not related to the subject at hand. See footnote seven.)
4) If it is a statement of a Companion, because the statements of the Companions are not the main purpose of this book, and therefore Imam Bukhari did not feel the need to always provide the chains leading up to their statements
5) Someone in the chain did not possess the required standard of exactitude in his narrations, or may be a trustworthy narrator but not according to the high standard that Imam Bukhari set for his Sahih book in particular.
Of the five reasons above, none of the first four could apply to the ‘Hadith of Instruments,’ which means that only the fifth reason applies to it: that one of its narrators did not meet the standards set by Imam Bukhari for this collection, which meant that the hadith chain as a whole falls below the standard set for this work.
In his great commentary Fath al-Bari, Ibn Hajar wrote specifically about the places where Imam Bukhari used the expression “said to us” (qala) rather than “narrated to us” for someone who was his own teacher, just like with the ‘Hadith of Instruments’:
وَقد ادّعى بن مَنْدَهْ أَنَّ كُلَّ مَا يَقُولُ الْبُخَارِيُّ فِيهِ قَالَ لِي فَهِيَ إِجَازَةٌ وَهِيَ دَعْوَى مَرْدُودَةٌ بِدَلِيلِ أَنِّي اسْتَقْرَيْتُ كَثِيرًا مِنَ الْمَوَاضِعِ الَّتِي يَقُول فِيهَا فِي الْجَامِعُ قَالَ لِي فَوَجَدْتُهُ فِي غَيْرِ الْجَامِعِ يَقُولُ فِيهَا حَدَّثَنَا وَالْبُخَارِيُّ لَا يَسْتَجِيزُ فِي الْإِجَازَةِ إِطْلَاقَ التَّحْدِيثِ فَدَلَّ عَلَى أَنَّهَا عِنْدَهُ مِنَ الْمَسْمُوعِ لَكِنْ سَبَبَ اسْتِعْمَالِهِ لِهَذِهِ الصِّيغَةِ لِيُفَرِّقَ بَيْنَ مَا يَبْلُغُ شَرْطَهُ وَمَا لَا يبلغ وَالله أعلم
I have studied a great number of places where [al-Bukhari] said in al-Jami’ (i.e. Sahih al-Bukhari), “Person X said to me,” and found that in another of his works he said “Person X narrated to me.” Al-Bukhari does not allow someone to use the words “he narrated to me” by someone who received a hadith (solely) via an ijaza (written or oral authorisation to narrate something), and this is proof that he heard these traditions from these people. However, the reason he used this expression (in the Sahih) is to distinguish between that which meets his standard of authenticity and that which does not, and Allah knows best.
So why would Imam Bukhari include such hadiths in his collections that do not meet the criteria of authenticity he chose for his work? When he did that, it was as a supplementary hadith used as a shahid (or witness) to another hadith of similar wording or meaning. The first hadith quoted would have been of al-Bukhari’s requisite standard of authenticity, and the shahid is added for extra support or for added information, even though it is not as strong. Therefore the first hadith would have been a musnad sahih hadith, as intended by al-Bukhari, and the shahid is a supplementary hadith, not to be taken on its own.
For example, elsewhere in Sahih al-Bukhari, Imam al-Bukhari narrated a hadith on the authority of the Companion Anas ibn Malik, starting the chain with “Qutayba ibn Sa’eed narrated to us… (until the end of the chain).” Imam Bukhari then followed it with a shahid, and wrote, “and Muslim [ibn Ibrahim] said to us: Aban [ibn Yazid al-Attar] narrated to us…(until the end of the chain).” Here for this shahid he used the expression “qala” rather than “haddathana.” The reason is because there is someone weak in the chain, who is Aban. Ibn Hajar wrote in his commentary,
Al-Bukhari does not quote any hadiths from [Aban] except as a shahid, and I have not seen any other fully-connected chain in this work that includes him. This is similar to the case of Hammad ibn Salama. Al-Bukhari said in the Chapter of Riqaq: “Abu al-Walid said to us: Hammad ibn Salama narrated to us…” This expression, “he said to us” is only used by al-Bukhari – according to what has been deduced from studying his work – mostly when narrating a shahid, and sometimes when narrating a statement of a Companion.
As we can see from the above examples, al-Bukhari would begin a hadith by saying that his own teacher “said” to him rather than “narrated” to him when someone else in the chain provided by the teacher is unacceptable to al-Bukhari, but there is some benefit or another for which al-Bukhari chose to include this chain in his work. However, in other works like his al-Tarikh al-Kabir, al-Bukhari would say that this teacher “narrated” this very same hadith to him because he did not set the same criteria of authenticity for al-Tarikh al-Kabir. Therefore, whenever al-Bukhari says in his Sahih that his teacher “said” a hadith to him, it is to alert the reader to the fact that this hadith does not meet the standards of this work and is there for another purpose.
The ‘Hadith of Instruments’
There is a chapter in Sahih al-Bukhari called the Chapter of Drinks (Kitab al-Ashriba). This chapter begins with four sections on the theme of what the prohibition of ‘khamr’ in the Qur’an covers. The reason Imam Bukhari included these sections is because some scholars at the time, especially in the Iraqi city of Kufa, argued that only the wine made of grapes was called ‘khamr.’ Therefore they argued that alcoholic drinks made from dates, honey, fruits, rice, or anything else were not prohibited, or that only an amount of them that led to intoxication was prohibited, but one could drink of them an amount that did not lead to intoxication. Imam Bukhari therefore included the following sections:
– Section on khamr Made of Grapes
– Section on khamr Made of Dates
– Section on khamr Made of Honey
– Section on khamr Being any Drink that Clouds (yukhamir) the Mind
In this last section, Imam Bukhari quoted a statement of Umar (may Allah be pleased with him) stating that when the prohibition of khamr was revealed, at that time the types of khamr known to the Prophet’s community were made of: Grapes, Dates, Wheat, Barley and Honey. However, Umar repeated three times, khamr was whatever clouded the mind, even if it was made from something other than the above five. The reason Imam Bukhari included all these sections was as a rebuttal to those early scholars who legalized date wine and other types of non-grape wine. All of these hadiths were made to establish legal rulings, and therefore needed to be rigorously authenticated, which they were. After this, to further drive his point, Imam Bukhari added one final section:
– Section on Those Who Legalize khamr Through Calling it by a Different Name
Here Imam Bukhari referred in the title of this section to a hadith that he did not even narrate in his Sahih, except through his paraphrased quote in the section title. It is in reality a different narration of the same ‘Hadith of Instruments’ that will follow it, but in different wording. This is something that Imam Bukhari has done a few times in his Sahih. The hadith in question goes (according to the wording in Bukhari’s al-Tarikh al-Kabir):
ليشربن نَاسٌ مِنْ أُمَّتِي الْخَمْرَ يُسَمُّونَهَا بِغَيْرِ اسْمِهَا يُضْرَبُ على رؤسهم بِالْمَعَازِفِ وَالْقَيْنَاتِ يَخْسِفُ اللَّهُ بِهِمُ الأَرْضَ وَيَجْعَلُ مِنْهُمُ الْقِرَدَةَ وَالْخَنَازِيرَ
People from my ummah will drink khamr, calling it by a different name, while singing girls play instruments over their heads. Allah will make the earth swallow them and make monkeys and pigs out of some of them.
Following this hadith which was paraphrased as a section title, Imam Bukhari quoted what we have so far been referring to as the ‘Hadith of Instruments.’ He wrote, as a continuation of the hadiths above:
وقال هشام بن عمار حدثنا صدقة بن خالد حدثنا عبد الرحمن بن يزيد بن جابر حدثنا عطية بن قيس الكلابي حدثنا عبد الرحمن بن غنم الأشعري قال حدثني أبو عامر أو أبو مالك الأشعري والله ما كذبني سمع النبي صلى الله عليه وسلم يقول ليكونن من أمتي أقوام يستحلون الحر والحرير والخمر والمعازف ولينزلن أقوام إلى جنب علم يروح عليهم بسارحة لهم يأتيهم يعني الفقير لحاجة فيقولون ارجع إلينا غدا فيبيتهم الله ويضع العلم ويمسخ آخرين قردة وخنازير إلى يوم القيامة
And Hisham ibn Ammar said to us: Sadaqa ibn Khalid narrated to us: Abd al-Rahman ibn Yazid ibn Jabir narrated to us: Atiyya ibn Qays al-Kilabi narrated to us: Abd al-Rahman ibn Ghanm al-Ash’ari narrated to us: he said: Abu Amir or Abu Malik al-Ash’ari narrated to us – by Allah he did not lie to me – that he heard the Prophet (may Allah exalt him and grant him peace) saying:
“From among my ummah there will be people who will consider the following to be lawful: illegal sexual intercourse (al-hir الحر), silk, khamr, and musical instruments. Some of them will stay near the side of a mountain and in the evening their shepherd will come to them requesting something he needs. They will say, ‘come back to us tomorrow.’ However, during that night Allah will destroy them and let the top of the mountain fall on them. Other people will be transformed into monkeys and pigs and remain thus until the Day of Resurrection.”
The first thing to notice about the hadith above is that it begins with “and” because it is a follow on to the hadiths above. In fact, it is only added to them as a ‘shahid.’ We can deduce this from the second noticeable thing, which is that Imam Bukhari did not start the narration with “Hisham ibn Ammar narrated to us,” but rather with “Hisham ibn Ammar said to us.” This is Imam Bukhari’s flag for us to realize that this hadith does not meet his criteria of authenticity for this work, and that it is only to function as a supplementary shahid following on from the previous hadiths. The function of this hadith is to drive home the point that those scholars who legalize date wine or any other type of non-grape wine are simply giving khamr a different name in order to classify it as lawful. They are going against the hadith that every drink that clouds the mind is classified as khamr.
Furthermore, the function of this hadith is not to set a new legal ruling, but to warn of dire punishment for those who legalize other types of wine and drink them. Imam Bukhari favored the most prestigious chains of transmission for setting new legal rulings, and used less prestigious chains (that are still rigorously authenticated, but might not have as many great hadith authorities in them) for hadiths about rewards, punishments, and other non-legal issues. If Imam Bukhari thought that this hadith was authentic enough to create a new legal ruling, which is the prohibition of musical instruments, he would have created a new section or chapter somewhere in Sahih Bukhari dedicated to the legality of musical instruments and placed the hadith there as his proof. This was the way of Imam Bukhari. There are 7,563 hadiths in Sahih Bukhari, but more than 3,000 of them are repetitions. The reason behind such repetitions is that Imam Bukhari repeated previously quoted hadiths in new sections of relevance, in order to support different legal rulings. He sometimes repeated a whole hadith or only took the relevant excerpt from it for the different section. Therefore, we could be absolutely sure that if Imam Bukhari thought this hadith conclusively showed the prohibition of instruments, he would have created a dedicated section to that and placed it, or just part of it, there. However, he did not, because this hadith does not meet his criteria for standing on its own in his book, let alone for setting a new legal ruling. Therefore Imam Bukhari took the rare step of leaving this hadith as mu’allaq even though he narrated it with its full chain, including his own teacher. By using the expression “said” instead of “narrated,” he was “leaving it hanging.”
That is why great scholars of hadith noted that this hadith was left as mu’allaq. The great hadith scholar al-Mizzi mentioned this in Tuhfat al-Ashraf and Tahdhib al-Kamal, mentioning in this latter work that it is the only hadith Imam Bukhari used which included the narrator Atiyya ibn Qays and that he only used it as a shahid. Al-Mizzi therefore marked Atiyya ibn Qays with the letters [Kh T] to show that he was used by al-Bukhari only as a shahid, rather than [Kh] which would mean that al-Bukhari used him in a ‘proper’ hadith from his Sahih (see footnote 10). Imam al-Dhahabi wrote: “Al-Bukhari narrated it from Hisham ta’liqan (as a mu’allaq). He said, ‘and Hisham said…’ (wa qala Hisham).”
Problems with the Hadith
This hadith has several problems, both in regards to its chain of transmission and the wording itself. Let us begin with the wording (matn). Abu Dawud narrated this same hadith with the same exact chain down from the Messenger of Allah (may Allah exalt him and send him greetings of peace), four steps downward until Abd al-Rahman ibn Yazid ibn Jabir. This means the only difference between the chain of al-Bukhari and the chain of Abu Dawud is in their immediate teacher and their teacher’s teacher.
Al-Bukhari has: Hisham from Sadaqa
Abu Dawud has: Abd al-Wahhab ibn Najda from Bishr ibn Bakr.
After that the chain is identical. However, the hadith in the text of Abu Dawud states,
From among my ummah there will be people who will consider the following to be lawful: silk gauze (al-khazz الخز) and silk.
Here instead of al-hir, a word which literally means private parts and refers to sexual intercourse, there is al-khazz, a type of raw silk gauze (in fact the word gauze comes from the Arabic khazz). In Arabic the two alternative words look exactly the same, the difference being in whether there is a dot over the letters. For this reason, Abu Dawud narrated this hadith in the Chapter of Clothing, under the Section on What Was Narrated Regarding Khazz. In fact, Ibn al-Athir holds that the more widely known narration of this hadith is with the word khazz, not al-hir like al-Bukhari’s narration. This narration of Abu Dawud does not mention musical instruments at all. Abu Dawud then comments on the hadith by stating that more than twenty Companions were known to wear khazz, implying therefore that it was not unlawful like silk. Abu Dawud therefore was implying that this hadith is not definitive as a source of law.
Another version of this hadith was narrated by al-Bukhari in al-Tarikh al-Kabir and others. It is the one paraphrased as a section title in Sahih al-Bukhari. Though we have quoted the hadith above, we can repeat it here:
People from my ummah will drink khamr, calling it by a different name, while singing girls play instruments over their heads. Allah will make the earth swallow them and make monkeys and pigs out of some of them.
This is how it was also narrated by Ibn Majah, Ibn Hibban, al-Bayhaqi (in his Sunan and in Shu’ab al-Iman), Ibn Abi Shayba, al-Tabarani (in al-Kabir), and Ibn Asakir in Tarikh Dimashq.
As the great Imam Ibn Hazm al-Zahiri wrote:
“There is nothing in the wording of this hadith that this warning is tied to listening to musical instruments, nor that it is about singing girls. The apparent meaning of the hadith is that the warning is for their making khamr lawful by giving it a different name. We do not base our religion on conjecture.” 
We see that for the very same hadith we have very different wordings. In one there is hir while another has khazz. In one there is mention of musical instruments and in another there is none at all. In one musical instruments are being declared lawful (meaning that they are in fact unlawful), while in another, the instruments are not the problem, they are simply part of the background scene, and the problem is in khamr being declared lawful by being given a different name.
That is why Shaykh Dr. Akram Nadwi, a great expert on hadith science and on Sahih al-Bukhari, said that every part of this hadith has been criticized by the hadith experts except the mention of alcoholic drinks. That is the only ‘reliable’ content of this hadith, and that is why Imam Bukhari used this hadith, only for this reliable content. Had the rest of the hadith’s content been reliable, he would have extracted from it more points of fiqh through new section headings elsewhere in the book, such as a section on musical instruments. That is why Shaykh Akram Nadwi also states that Imam Bukhari never intended to use this hadith as evidence against musical instruments, and that those who do use this hadith for that reason are either misleading people, or are ignorant of the fact that it is mu’allaq and that Imam Bukhari found it problematic.
We have seen that the hadith has major problems and uncertainity in terms of its matn or text. Not only that, all the different versions have problems with their chains of transmission too.
Problems with the chains of transmitters.
As for the version in al-Tarikh al-Kabir and all the other hadith works just mentioned above, it was narrated on the authority of Malik ibn Abi Maryam. This is problematic because Ibn Hazm wrote, “It is not known who he is,” while al-Dhahabi wrote, “He is not known.” The one who narrated this hadith in turn from Malik is Hatim ibn Hurayth, who was not very well known and narrated very few traditions, so that the great Yahya ibn Ma’in said, “I do not know him,” while those who did know him described him as truthful and that he narrated acceptable traditions. Al-Jurjani wrote, “I hope that there is nothing wrong with him.” Because of such major uncertainties, especially with one transmitter (Malik ibn Abi Maryam) being majhul (unknown), al-Bukhari did not narrate the hadith in his Sahih, but referred to it in a section heading. Even when he narrated it in al-Tarikh al-Kabir, he again used the expression “qala” (said) just like he did for the other version in his Sahih. He wrote, “Sulayman ibn Abd al-Rahman said it to me.”  This immediate teacher of al-Bukhari was himself truthful but was known to make mistakes, and more than that, he was well-known for frequently narrating from weak narrators. Abu Hatim al-Razi said: “He was truthful but he narrated from the weak and unknown narrators more than anyone else.” Ibn Hibban said: “His hadiths are only taken into account when he narrates from well-known trustworthy people,” which is what Yahya ibn Ma’in also said. Ibn Hajar said: “He is truthful but makes mistakes,” and “al-Bukhari only narrated very few hadiths from him.” Al-Dhahabi and al-Daraqutni both said that he himself was trustworthy but narrated from many people who are to be rejected.
As for the version that al-Bukhari used in his Sahih, the first problem with the chain is that it is mu’allaq, meaning that al-Bukhari worded it in a way that we do not know if he actually heard it from the person above him in the chain. However, we do know from other places in al-Bukhari that he did narrate hadiths from this Hisham, and in one instance even narrated from Hisham a hadith from Sadaqa, the next person in the chain. This tells us that al-Bukhari very likely did hear this hadith from his teacher Hisham, but we cannot be certain. Standing alone like that, based on internal evidence from Sahih Bukhari, we cannot be certain that al-Bukhari heard this hadith in particular from this teacher, which means we cannot rely upon it. That is why Ibn Hazm said that this hadith was disconnected between al-Bukhari and Sadaqa. Some scholars criticised Ibn Hazm for saying this, because we know that al-Bukhari met Hisham and took hadiths from him, including a hadith from Sadaqa, but Ibn Hazm is right in that this does not prove that this particular hadith of Hisham from Sadaqa was also heard by al-Bukhari from Hisham, and therefore this hadith remains disconnected and cannot be relied upon.
If we are to assume that al-Bukhari did hear this hadith from Hisham, then the question would be why he left it as mu’allaq from his own teacher, a usual indication of a weakness in the chain as mentioned above from Ibn Hajar’s analysis of the Sahih Bukhari as a whole and all the mu’allaq hadiths. The answer is that the weakness is in the Successor Atiyya ibn Qays. In this version, Atiyya could not remember which Companion he took the hadith from, naming two of them, saying he could not remember which one it was. The fact that we do not know which Companion narrated the hadith is not a problem in itself because all Companions are considered trustworthy. However, this indicates that someone in the chain does not have an accurate memory. Ibn Hajar identified Atiyya ibn Qays as the one with the doubt because the other person to narrate from the Companion was Malik ibn Abi Maryam (discussed in the paragraph above), and in the version with Malik there is no doubt about which Companion narrated the hadith. Atiyya was a very pious man from the generation of the Successors and a great teacher of Qur’an recitation. He was truthful and pious, but not a trustworthy narrator, because his memory and exactness did not meet the standards of the top hadith critics. That is why he was classified as “salih al-hadith” (his hadiths are acceptable/valid) by Abu Hatim al-Razi, and “la ba’sa bih” (he is ok) by al-Bazzar. Abu Hatim al-Razi’s son wrote in al-Jarh wal-Ta’dil:
When it is said that someone is fully trustworthy (thiqa) or is exact in his narrations (mutqin thabt), then it is someone whose narrations are taken as proof.
When it is said that someone is truthful or that he is ok, then his hadiths are written down and studied. This is the second degree.
When it is said that someone is a “shaykh” then he is in the third degree. His hadiths are written down and studied, but it is below the second degree.
When it it said that someone is “salih al-hadith” then his hadiths are only written down to be taken into consideration.
Therefore al-Bazzar placed him in the third degree and Abu Hatim al-Razi placed him in the fourth, and Imam Abu Hatim al-Razi is more knowledgeable than al-Bazzar. Ibn Abi Hatim then continued to quote the great hadith expert and student of Imam Malik, Abd al-Rahman ibn Mahdi, as using the expression “salih al-hadith” for people who were truthful but weak and made mistakes. Al-Albani said, commenting on the above passage of Ibn Abi Hatim:
“This is a clear text that their use of “salih al-hadith” is similar to their use of “layyin al-hadith” (weak -literally: soft- in their narrations), meaning that their hadiths are only written down as a shahid and to be taken into consideration. This means that the narrations of such a person cannot be used as evidence. Therefore for Abu Hatim (al-Razi) this is an expression of criticism (jarh) and not praise (ta’dil).”
That is why, as we have seen above, al-Mizzi also classified Atiyya ibn Qays not as someone that al-Bukhari accepted, but as someone that al-Bukhari only accepted as a shahid, in a mu’allaq hadith. That is a very different classification and shows that al-Bukhari’s use of a hadith by him does not mean that al-Bukhari accepted him on his own or accepted his hadiths as proofs.
We have seen that Atiyya ibn Qays (may Allah be pleased with him) made the mistake of forgetting which Companion he heard the hadith from. Furthermore, his narration is the only one in which musical instruments are mentioned as forbidden, whereas in all the other more widespread and well-known narrations, musical instruments are not mentioned as forbidden. They are only mentioned in the background scene. It therefore seems that Atiyya ibn Qays mistakenly mentioned musical instruments along with khamr as things that are unlawful but made lawful by these people about whom the hadith fortells, rather than with the things mentioned later in the hadith in the background scene.
Because of the major problems with this version of the hadith – the only version criticizing musical instruments – it cannot be found in Sahih Muslim, any of the major Sunan works, or the Musnad of imam Ahmad ibn Hanbal. In other words, this hadith’s many problems made it unworthy of entry into any well known hadith work, and that is why it can only be found as a mu’allaq in Sahih Bukhari and no where else. Had Imam Bukhari not wanted to use only one part of it to make a point, most people would never have known that such a narration even existed.
Why Did al-Bukhari Include this Hadith in his Sahih?
We have mentioned above that part of Bukhari’s Chapter on Drinks is dedicated to a refutation of the position of his enemies from the school of the Ahl al-Ra’y in Kufa, that any alcoholic beverage other than grape wine does not classify as khamr and is therefore allowed. He thus dedicated four sections to refute them, quoting authentic hadiths to establish that wine made from dates, honey, or any (alcoholic) drink that clouds the mind, are unlawful. Those were the hadiths that Bukhari relied upon, which were fully connected and authentic. He followed that with a paraphrase of the ‘Hadith of the Instruments,’ the version that in fact does not say that instruments are unlawful in any way, but says that there will be people who give different names to khamr to make it legal. This is what al-Bukhari was saying those people who allowed other alcoholic drinks were doing, However, he could not place the relevant tradition in this section because one of its narrators was completely unknown (majhul), so he had to suffice himself by referring to it. Then al-Bukhari quoted the other version of the same hadith which did not say that these people gave alcoholic drinks a different name, but was good enough in that it said that they will consider alcoholic drinks to be lawful. This hadith at least did not have any unknown people in its chain, and its problem was the existence of one narrator, Atiyya ibn Qays, who made mistakes and did not have a good enough memory, so al-Bukhari could use it as a shahid. Al-Bukhari pointed out that it was only being used as a shahid by saying “Hisham said” rather than “narrated to us.” It is as if he included this hadith as a warning to his enemies the Ahl al-Ra’y jurists of Kufa, while at the same time he left a sign there for his own compatriots, the hadith experts, that this hadith cannot be relied upon on its own – it is there only as a supplement to make his point and not to establish any legal rulings. Imam Bukhari recognized that despite the weakness of both versions of the hadith, they both concurred on the prohibition of alcoholic drinks. Therefore he included them in this chapter. However, they did not concur on the prohibition of musical instruments, and neither one of them on its own, or both together, were enough to establish the prohibition of musical instruments as a ruling. That is why, even though al-Bukhari had more than 3,000 repititions in Sahih al-Bukhari, to use hadiths for as many relevant topics as possible, and to extract as much fiqh from them as possible, he never created a section or chapter on musical instruments based on these hadiths.
This distinction between hadiths used as the main part of Sahih al-Bukhari, those hadiths that were authentic and musnad, and those hadith which are only added as supplementary material, was well known to the hadith masters of the past. That is why they distinguished between hadiths in Sahih al-Bukhari that could be used as proof, and hadiths in Sahih al-Bukhari that could not, on their own, be used as proof. They used this distinction also in classifying narrators. There were narrators accepted by al-Bukhari in his Sahih for those musnad hadiths, and narrators only used by al-Bukhari in the same work for his mu’allaq hadiths and/or as a shahid. However, many people today are unaware of this distinction and they simply say: “Musical instruments are definitely unlawful because of the authentic hadith in Sahih al-Bukhari.” On the other hand, the great jurist and hadith expert Abu Bakr ibn al-Arabi said in Ahkam al-Qur’an that none of the hadiths quoted against musical instruments are in any way authentic  and Ibn Hazm, whom Ibn Taymiyya counted among a handful of the greatest hadith critics, said, “There is absolutely nothing authentic on this topic (i.e. musical instruments) at all. Everything about this topic is fabricated.”  Ibn Hazm wrote a treatise dedicated to showing the faults in all the hadiths against musical instruments and singing, and showing all the authentic hadiths proving their permissibility. The great hadith master and jurist Ibn Abd al-Barr studied this treatise and said, “I have not found anything to add to it or remove from it.” 
 See the definition of a musnad hadith in Ibn Hajar al-Asqalani, Nuzhat al-nadhar, Cairo: Dar al-Salam, 2006, pp. 190-1.
 Ibn Hajar al-Asqalani, Fath al-Bari, vol. 1, p. 469 (From Ibn Hajar’s introduction to Fath al-Bari).
 Ibn Hajar, Taghliq al-Ta’liq, vol. 2, p. 8. (The first volume is the editor’s study of Ibn Hajar’s work); Ibn Hajar, Nuzhat al-nadhar, p. 119.
 If such an author is known as someone who commits tadlees (hides the intermediaries between himself and someone else), then the hadith is not classified as mu’allaq, but as being afflicted with tadlees. However, if the author would never commit tadlees, like the great imam Bukhari, then the hadith is classified as mu’allaq, whether in fact he did hear it from that person above directly or through intermediaries. See Ibn Hajar, Nuzhat al-nadhar, p. 119.
 Ibn Hajar, Taghliq al-Ta’liq, vol. 2, p. 8.
 Ibn Hajar, Fath al-Bari, vol. 1, p. 156 (Kitab al-Ilm, Bab ma yudhkar fi’l-munawala).
 Ibn Hajar, Fath al-Bari, vol. 5, p. 3 (Kitab al-Muzara’a, Bab fadl al-zar’ wal-ghars idha ukila minh). Ibn Hajar then explained why al-Bukhari quoted this other chain as a shahid without quoting the text of the hadith. The reason is because in the first chain, the Successor Qatada is quoted as saying “on the authority of Anas,” which leaves the possibility that Qatada did not hear it directly from Anas. The supplementary chain quotes Qatada as saying “Anas narrated to us,” and it is for this benefit that al-Bukhari appended this other chain after narrating the first hadith.
 Al-Bukhari, al-Tarikh al-Kabir, vol. 1, p. 305.
 Jamal al-Din al-Mizzi, Tuhfat al-Ashraf, vol. 9, p. 282.
 Jamal al-Din al-Mizzi, Tahdhib al-Kamal, vol. 1, p. 149, where he states that hadiths marked as ‘Kh’ [خ] are hadiths from Bukhari’s Sahih, but hadiths marked as “Kh T” [خت] means they are used by al-Bukhari in his Sahih only as a shahid in the form of a mu’allaq (the T stands for ta’leeqan or mu’allaq). Al-Mizzi marked the ‘Hadith of Instruments’ with “Kh T” [خت] in Tuhfat al-Ashraf and in Tahdhib al-Kamal. In his entry on Atiyya ibn Qays, one of the narrators in the chain of the ‘Hadith of Instruments,’ al-Mizzi wrote that “al-Bukhari used only one hadith narrated through him, as a shahid,’ before quoting the ‘Hadith of Instruments.’ See al-Mizzi, Tahdhib al-Kamal, vol. 20, pp. 153, 156.
 Al-Dhahabi, Siyar A’lam al-Nubala’ , entry on Abu Musa al-Madini
 Ibn Hazm, al-Muhalla, vol. 7, p. 562.
 Al-Bukhari, al-Tarikh al-Kabir, vol. 1, p. 305.
 Ibn Hajar, Fath al-Bari, vol. 10, p. 55. The right companion is Abu Malik al-Ash’ari.
 Ibn Abi Hatim al-Razi, al-Jarh wa’l-Ta’dil, Beirut: Dar Ihya al-Turath al-Arabi, vol. 2, p. 37.
 Ibn Abi Hatim al-Razi, al-Jarh wa’l-Ta’dil, Beirut: Dar Ihya al-Turath al-Arabi, vol. 2, p. 37.
 Al-Albani, al-Silsila al-Da’ifa, vol. 3, p. 112.
 Ibn al-Arabi, Ahkam al-Qur’an, vol. 3, p. 526.
 Ibn Hazm, al-Muhalla, vol. 7, p. 565. See Ibn Taymiyya, al-Fatawa al-Kubra, vol. 3, p. 283, where he lists examples of the “great imams who are experts in the hidden faults of hadiths and their fiqh, like Imam Ahmad ibn Hanbal and al-Bukhari and others, and Abu Ubayd and Abu Muhammad Ibn Hazm, and others…”
 Ibn Hazm, Risala fil-Ghina’ al-Mulhi, in Rasa’il Ibn Hazm al-Andalusi (ed. Ihsan Abbas), vol. 1, p. 439.