The people involved in the transmission of a hadeeth constitute its isnad.  The isnad informs us about the hadeeth’s source, and this information later became an essential part of the hadeeth (Azami 31).  Abdullah b. Al-Mubarak, one of the teachers of al-Bukhari, is reported to have said, “The isnad is part of the religion: had it not been for the isnad, whoever wished to would have said whatever he liked” (Hasan 11)[1].  There is some indication that the isnad was used before the first tribulation, though it was not until the end of the first century of the Hijrah that it was fully developed (Azami 33).  (However, John Burton in his An Introduction to the Hadith says that the isnad did not yet exist in the first century) The other part of the hadeeth that actually contains the specific saying or action of the Prophet, may the mercy and blessings of God be upon him, is its matn or text.

For the classification of hadeeth, there are several broad categories, of which only seven will be very briefly discussed here.  The seven categories are classifications according to 1) the reference to a particular authority, 2) the links in the isnad, 3) the number of reporters involved in each stage of the isnad, 4) the technique used in reporting the hadeeth, 5) the nature of the isnadand matn, 6) a hidden defect found in the hadeeth’s isnad or matn, and 7) the reliability and memory of the reporters (Hasan 14-16).

The first category, classification according to the reference to a particular authority, pertains to whether it goes back to the Prophet, a Companion, or a Successor.  A marfu’ or “elevated” narration is one that back to the Prophet, and this is regarded as the best grade (Burton 112).  A mawqoof or “stopped” narration is one that goes back to a Companion, while a maqtu’ or “severed” narration is one that goes back to a Successor.  This classification is significant in that it differentiates between the Prophet’s sayings and actions and that of a Companion or Successor.

The second category, classification according to the links in the isnad, makes several different distinctions.  The musnad or “supported” hadeeth is the best out of the group as it contains no break in the chain of authorities reporting the hadeeth back to the Prophet (Burton 111).  The mursal or “unattached” hadeeth is one that contains a gap of one generation (according to both Azami and Hasan it is a hadeeth reported by a Successor who drops the Companion from whom he learned it in the isnad).  The munqati’ or “broken” hadeeth is one which is missing a link closer to the traditionalist reporting it (i.e., before the Successor).  This applies even if there appears to be no break in the chain, if it is known that one of the reporters could not have heard hadeeth from the immediate authority given in the isnad, even if they are contemporaries.  The term munqati’also is used by some scholars to refer to a hadeeth in which a reporter does not name his authority and instead says, “a man narrated to me” (Hasan 22).  A hadeeth is mu’dal or “perplexing” if more than one consecutive reporter is missing in the isnad.  If the isnad is dropped altogether and the reporter directly quotes the Prophet, then the hadeeth is considered mu’allaq or “hanging” (Hassan 22).

Within the third category, hadeeth are classified according to how many reporters are in each stage of the isnad, i.e. in each generation of reporters.  The two main classifications are mutawatir(“consecutive”) and ahad (“single”), though ahad is further divided into many subdivisions, among them ghareeb (“scarce” or “strange”), ‘azeez (“rare” or “strong”), and mash’hoor (“famous”).  A mutawatir hadeeth is one that is reported by a large number of people whose agreement upon a lie is not reasonably possible and in which the possibility of coincidence is negligible.  The minimum number of required reporters differs among the scholars of hadeeth, and ranges from four to several hundred (Azami 43).  The hadeeth may be mutawatir in either meaning or words, the former being the more common one.  Al-Ghazali stipulated that the hadeeth must be mutawatir in the beginning, middle, and last stages of its isnad (Hasan 30).  A hadeeth that is ahad is one whose number of reporters does not come near to that required for a mutawatirhadeeth.  A hadeeth is classified as ghareeb if at any stage (or every stage) in the isnad there is only one person reporting it.  A hadeeth is classified as ‘azeez if at every stage in the isnad there are at least two people reporting it.  If at least three people report a hadeeth in every stage of its isnad, then it is classified as mash’hoor, although the term is also applied to those hadeeth which start out as ghareeb or ‘azeez but then end up with a larger number of reporters (Hasan 32).

In the fourth category, hadeeth are classified according to manner in which they are reported.  As was mentioned earlier, there is a corresponding special term to denote a particular mode of learning or transmission when a student or scholar learned a hadeeth.  “Haddathana,” “akhbarana,” and “sami’tu” all indicate that the reporter personally heard the hadeeth from his own sheikh.  “‘An” and “qaala” are more vague and can signify either hearing from the sheikh in person or through someone else.  Actually, “‘an” is very inferior and can signify learning the hadeeth through any one of various modes of transmission (Azami 22).  A hadeeth can be labeled as weak due to the uncertainty caused by using the latter two terms, which respectively translate into “on the authority of” and “he said” (Hasan 33).  One who practices tadlees, “concealing”, reports from his sheikh that which he did not hear from him, or reports from a contemporary whom he never met.  This violates the principle that a hadeeth must be heard first-hand in order to be transmitted (Burton 112).  Another type of tadlees, which is considered the worst among them, is when a reliable scholar reports from a weak authority who is in turn reporting from a reliable scholar.  The person who is reporting this isnad may show that he heard it from his sheikh, but then omits the weak authority and simply uses the term “‘an” to link his sheikh with the next trustworthy one in the isnad (Hasan 34).

If throughout the isnad all the reporters (including the Prophet) use the same mode of transmission, repeat an additional statement or remark, or act in a particular way while narrating the hadeeth, then it is called musalsal (“uniformly-linked”).  This type of knowledge is useful for discounting the possibility of tadlees in a particular hadeeth (Hassan 35).

According to the fifth category, a hadeeth can also be classified with respect to the nature of its text and isnad.  According to Al-Shafi’i, if a hadeeth reported by a trustworthy person goes against the narration of someone more reliable than him, then the hadeeth is shadh or “irregular”.  According to Ibn Hajar, if a narration by a weak reporter contradicts an authentic hadeeth, then that hadeeth is classified as munkar (“denounced”), although some scholars would classify any hadeeth of a weak reporter as munkar.  A hadeeth could also be classified as munkar if its text contradicts general sayings of the Prophet.  If a hadeeth reported by a reliable person contains some additional information not narrated by other authentic sources, the addition is accepted so long as it doesn’t contradict them, and the addition is known as ziyadatu thiqah (“an addition by one trustworthy”).  However, if a reporter adds something to the hadeeth being narrated, then the hadeeth is classified as mudraj or “interpolated”.  If this occurs in a hadeeth, then it is usually in its text and often for the purpose of explaining a difficult word.  In a few examples this occurs in the isnad - a reporter takes a part of one isnad and adds it to another isnad.  A reporter found in the habit of intentional idraj or interpolation is generally considered a liar, although scholars are more lenient with those reporters who may do it to explain a difficult word (Hasan 37-39).

In the sixth category, hadeeth that contain hidden defects in their isnad or text are classified as ma’lool or mu’allal (“defective”).  This could be due to such things as classifying a hadeeth as musnad when it is actually mursal or attributing a hadeeth to a particular Companion when it really comes from another one.  In order to detect such defects, all the isnads of a hadeeth have to be collected and examined.  For example,

“Some scholars wrote works on which Successors heard hadeeth from which Companions.  From this information is it known that Al-Hasan Al-Basri did not meet Ali, although there is a slight chance that he may have seen him during his childhood in Madinah.  This is significant as many Sufi traditions are said to go back to Al-Hasan Al-Basri who is said to have reported directly from Ali.” (Hasan 42-43)

There can also be uncertainty about the isnad or text, in which case the hadeeth is classified as mudtarib (“shaky”).  This occurs if reporters disagree about some points in the isnad or text in such a way that no opinion prevails.  A hadeeth may be classified as maqloob (“changed” or “reversed”) if in the isnad a name was reversed (i.e., Ka’b b. Murra versus Murra b. Ka’b) or if the order of a sentence in the text is reversed (Azami 66).  This also applies to those hadeeth whose text has been given a different isnad or vice versa, or those in which a reporter’s name was replaced with another (Hasan 41-42).

The seventh and last category to be discussed here is classification according to the quality of the reporters, upon which the final verdict on a hadeeth critically depends.  Hadeeth reported by those known to be adil, hafiz, thabit, and thiqa are the highest ranked hadeeth and are classified as saheeh or “sound.”  For someone to be considered adil, he had to be a very pious Muslim, honest and truthful in all of his dealings.  Through careful comparison, verbal agreement found in the text of a hadeeth among various transmitters indicated who was the most accurate (thabit), the most reliable (thiqa), and who had the best memory (hafiz).  If any scholar falls less than this ideal in one or more categories, but he is not criticized, then the hadeeth reported by him are judged to be less sound, or hasan (“fair”).  If a reporter was known to have a weak memory or make mistakes due to carelessness, then his hadeeth are judged as da’eef (“weak”) (Burton 110-111).

Of course, there are other factors which play into the final verdict on a hadeeth, and in the words of Ibn Al-Salah, “A saheeh hadeeth is the one which has a continuous isnad, made up of reporters of trustworthy memory from similar authorities, and which is found to be free from any irregularities (i.e. in the text) or defects (i.e., in the isnad).”  According to Al-Tirmidhi a hasan hadeeth is “A hadeeth which is not shadhdh, nor contains a disparaged reporter in its isnad, and which is reported through more than one route of narration” (Hasan 44-46).  A hadeeth that doesn’t reach the requirements for a hasan hadeeth is classified as da’eef, and often this is due to discontinuity in the isnad.  It can also be classified as da’eef if one of the reporters does not have a good reputation for whatever reason, be it because of his making many mistakes or being dishonest.  If the defects are many and severe, then the hadeeth is closer to being classified as mawdu’ or fabricated.  According to Al-Dhahabi the mawdu’ hadeeth is the one whose text goes against established norms of the Prophet’s sayings or whose isnad contains a liar.  A hadeeth can also be established as mawdu’ due to “external evidence related to a discrepancy found in the dates or times of a particular incident” (Hasan 49).

In conclusion, the aforementioned classifications constitute only a fraction of the total number of classifications that exist.  The studies in hadeeth are very complex, and it seems that the scholars thought of every imaginable angle from which to analyze hadeeth.  All this was for the purpose of distinguishing between different types of narrations, especially for distinguishing the authentic from the inauthentic.