top of page

Knowledge as a Vritti

Knowl­edge plays a cru­cial role in all of our lives. We are not just speak­ing of knowl­edge of spe­cif­ic sub­jects that we learn in school, but also of knowl­edge gained from life expe­ri­ence, per­tain­ing to peo­ple, places and sit­u­a­tions. From sim­ple tasks like rec­og­niz­ing the face of a per­son we know, to mak­ing com­plex deci­sions in our pro­fes­sion­al and per­son­al life, it is knowl­edge that under­lies these process­es.

What exact­ly is knowl­edge? In cog­ni­tive psy­chol­o­gy, knowl­edge is defined as infor­ma­tion about the world that is stored in mem­o­ry; this infor­ma­tion must be capa­ble of being jus­ti­fied as true and be coher­ent. Inter­est­ing­ly, in the Yoga Sutras of Patan­jali, verse 1.6 talks about pra­mana, valid means of knowl­edge, as one of the five cat­e­gories of vrit­tis (mod­i­fi­ca­tions or fluc­tu­a­tions) of the mind-field, chit­ta. The sub­se­quent verse (1.7) men­tions the three modes of pra­mana, cor­rect means of knowl­edge:

  1. Pratyak­sha — direct per­cep­tion through one’s sens­es

  2. Anu­mana — infer­ence

  3. Aga­ma — tes­ti­mo­ny of accom­plished experts

Trans­la­tion: Direct per­cep­tion, infer­ence and revealed author­i­ty are the three cat­e­gories of the vrit­ti called valid proof (pra­mana)

Before going into more detail about the three means of pra­mana, let us look at how knowl­edge is rep­re­sent­ed in the brain. Infor­ma­tion that stands for an object, event or con­cept is phys­i­cal­ly rep­re­sent­ed as a net­work of neu­ronal con­nec­tions in par­tic­u­lar regions of the brain. (A neu­ron is the fun­da­men­tal unit of the brain and ner­vous sys­tem.)

For­mats of knowl­edge rep­re­sen­ta­tion in the brain

Research in the fields of psy­chol­o­gy and neu­ro­science points to the fact that the brain uses mul­ti­ple for­mats of rep­re­sen­ta­tion of knowl­edge: images, fea­ture detec­tors and sta­tis­ti­cal pat­terns in neur­al nets. Mul­ti­ple for­mats are required because the act of cog­ni­tion con­sists of numer­ous process­es.

Let’s under­stand these rep­re­sen­ta­tion­al for­mats with an exam­ple. Imag­ine that you are invit­ed to a birth­day par­ty and you are per­ceiv­ing the the table with the cake in front of you. When the eyes per­ceive the scene, the brain con­structs a visu­al image of it in the occip­i­tal cor­tex, locat­ed at the back of the head. Images are modal­i­ty-spe­cif­ic rep­re­sen­ta­tions; that is, they stand for infor­ma­tion received by a sin­gle sen­so­ry modal­i­ty, name­ly vision. Sim­i­lar­ly, infor­ma­tion com­ing from the sen­so­ry modal­i­ties of audi­tion, touch, taste and smell are rep­re­sent­ed as sense-spe­cif­ic mem­o­ries. As the image devel­ops, fea­ture detect­ing neu­rons extract mean­ing­ful fea­tures from the image such as the shape and colour of the cake, the tex­ture of the crumb, the frost­ing, the let­ters in edi­ble ink and so on. The fea­ture detec­tion hap­pens in cer­tain regions of the occip­i­tal, tem­po­ral and pari­etal lobes. Final­ly, a net­work of con­junc­tive neu­rons in the tem­po­ral lobes link the neu­rons active in form­ing the image, along with the neu­rons active in the fea­ture detec­tion. These con­junc­tive neu­rons rep­re­sent the sta­tis­ti­cal pat­tern that asso­ciate all the infor­ma­tion togeth­er to rep­re­sent the ‘birth­day cake’ (Fig­ure 1).

Fig­ure 1: For­mats of rep­re­sen­ta­tion in the brain. Neu­rons active in form­ing the image of the cake (in occip­i­tal lobes) and fea­ture detect­ing neu­rons (in occip­i­tal, tem­po­ral and pari­etal lobes) are linked togeth­er by con­junc­tive neu­rons in the tem­po­ral lobes to form a sta­tis­ti­cal pat­tern that rep­re­sent the ‘birth­day cake’. (Image cour­tesy: Smith and Koss­lyn, Cog­ni­tive Psy­chol­o­gy: Mind and Brain).

Sta­tis­ti­cal pat­terns: Knowl­edge rep­re­sen­ta­tions as sta­tis­ti­cal pat­terns in neur­al nets are becom­ing increas­ing­ly inter­est­ing to researchers. The neur­al net is a for­mat of rep­re­sen­ta­tion in which an object, say the birth­day cake, is rep­re­sent­ed as a sta­tis­ti­cal pat­tern such as 1100101000101. Bio­log­i­cal­ly, the ele­ments of a sta­tis­ti­cal pat­tern can be looked at as neu­rons (or pop­u­la­tions of neu­rons) that are “ON” or “OFF, i.e., they fire or do not fire. When a neu­ron is said to “fire”, it means that it emits an elec­tri­cal sig­nal called action poten­tial, which is impor­tant for com­mu­ni­ca­tion between the neu­rons. The 1’s in the pat­tern indi­cate neu­rons that fire and the 0’s indi­cate neu­rons that do not fire.

These rep­re­sen­ta­tions lay the ground­work for knowl­edge. Once the brain estab­lish­es mem­o­ries that car­ry infor­ma­tion about the world, all kinds of sophis­ti­cat­ed cog­ni­tive abil­i­ties become pos­si­ble. This par­tic­u­lar insight from cog­ni­tive neu­ro­science shows the impor­tance of stored rep­re­sen­ta­tions in mem­o­ry for high­er cog­ni­tive abil­i­ties such as rea­son­ing, deduc­tion and deep­er under­stand­ing. Per­haps this is the rea­son why the Indi­an edu­ca­tion sys­tem empha­sizes mem­o­riza­tion and rote learn­ing, as it forms the basis for under­stand­ing. This is a tan­gent to the main top­ic of this arti­cle, but worth men­tion­ing nev­er­the­less.


In the Vyasa Bhashya for Verse 1.7, it is explained that the mind-field is drawn towards and coloured by the objects of expe­ri­ence through the sens­es. The mind-field thus takes the form of a vrit­ti (mod­i­fi­ca­tion) whose con­tent is the sensed object; this vrit­ti is called pratyak­sha, the means of knowl­edge through direct per­cep­tion. It is said that pratyak­sha “deter­mines pri­mar­i­ly the spe­cif­ic in a mat­ter con­sist­ing of the gen­er­al and spe­cif­ic”. In oth­er words, dur­ing pratyak­sha, it is the spe­cif­ic object that is per­ceived and not the gen­er­al con­cept. For exam­ple, if one per­ceives a cat walk­ing by, it is the spe­cif­ic cat, with a spe­cif­ic colour, size, fur den­si­ty and so on per­ceived at that moment in time through the sens­es. It is not the gen­er­al con­cept of cats that we have in our minds, that is, an image of an aver­age cat that we have come across many a time.

It is fur­ther explained that dur­ing pratyak­sha, the deter­mi­na­tive process of the bud­dhi (intel­li­gence) helps in iden­ti­fy­ing spe­cif­ic attrib­ut­es that are unique to the object, dis­tin­guish­ing them from gen­er­al or com­mon attrib­ut­es shared with oth­er objects. For exam­ple, one might look at a but­ter­fly and think of the gen­er­al cat­e­go­ry of winged insects that includes but­ter­flies and moths, but per­ceiv­ing the bright­ly coloured wings and large size, one specif­i­cal­ly deter­mines it to be a but­ter­fly.

Frontal lobes: Mod­ern neu­roimag­ing research shows that the brain’s frontal areas, which are wide­ly asso­ci­at­ed with exec­u­tive func­tions and intel­li­gence, are active while cat­e­go­riz­ing objects using such spe­cif­ic attrib­ut­es and def­i­n­i­tions.

Vyasa’s com­men­tary also states that it is Purusha, the pure con­scious­ness prin­ci­ple, that is the cause of buddhi’s accu­rate appre­hen­sion. It is Purusha that gives a sem­blance of con­scious­ness to the process of per­cep­tion and the resul­tant vrit­ti.


The Vyasa Bhashya describes anu­mana or infer­ence as the vrit­ti deter­min­ing pri­mar­i­ly the gen­er­al; it is the invari­ant asso­ci­a­tion between an object and its attribute or com­mon­al­i­ty between two objects that helps to pro­duce a con­clu­sion. The clas­sic illus­tra­tion giv­en is the asso­ci­a­tion between fire and smoke. Due to this asso­ci­a­tion, when one sees smoke, one infers that there is fire.

It is clear­ly explained that anu­mana is the vrit­ti that deter­mines pri­mar­i­ly the gen­er­al. For exam­ple, you know that birth­day par­ties are asso­ci­at­ed with cakes. If you see an unopened box on the table at a birth­day par­ty, you know that it is a cake even though you haven’t actu­al­ly seen it. The cake that appears in the mind is a ‘gen­er­al’ con­cept of cake and not the spe­cif­ic cake that is present in the box.

Cat­e­go­ry knowl­edge: Cog­ni­tive neu­ro­science refers to this process of asso­ci­a­tion of objects with sim­i­lar fea­tures as “cat­e­go­riza­tion”. Objects with com­mon attrib­ut­es or prop­er­ties can be grouped into cat­e­gories. Cat­e­go­ry knowl­edge devel­op­ment in a two step process: first, rep­re­sen­ta­tions of a indi­vid­ual mem­bers of a cat­e­go­ry are estab­lished; sec­ond, those rep­re­sen­ta­tions are inte­grat­ed. For exam­ple, we have seen dif­fer­ent types of cakes in our life. Each indi­vid­ual cake pro­duces a unique sta­tis­ti­cal pat­tern of 0’s and 1’s. Although the sta­tis­ti­cal pat­terns are unique for each cake, they are sim­i­lar because the actu­al cakes have com­mon fea­tures. In the sta­tis­ti­cal pat­terns, cer­tain con­junc­tive units are com­mon to all rep­re­sen­ta­tions (shown shad­ed in Fig­ure 2a). This shared sta­tis­ti­cal pat­tern estab­lish­es the rep­re­sen­ta­tion of the cat­e­go­ry, that is, the ‘gen­er­al con­cept of cake’, not just a spe­cif­ic cake (Fig­ure 2b). When some­one men­tions the word “cake”, the shared sta­tis­ti­cal pat­tern becomes active, even though a cake is not phys­i­cal­ly present, and pro­duces a men­tal sim­u­la­tion of a cake that is rough­ly the aver­age of pre­vi­ous­ly expe­ri­enced cakes (Fig­ure 2c).

Fig­ure 2. Rep­re­sen­ta­tion of cat­e­go­ry knowl­edge (a) Each indi­vid­ual cake has a unique sta­tis­ti­cal pat­tern. There are some com­mon units, reflect­ing the sim­i­lar­i­ties of actu­al cakes. (b) The shared units of the sta­tis­ti­cal pat­tern estab­lish the rep­re­sen­ta­tion of the ‘gen­er­al con­cept’ of cakes. © The word ‘cake’ acti­vates the shared sta­tis­ti­cal pat­tern and pro­duces a men­tal sim­u­la­tion of a cake that is rough­ly the aver­age of pre­vi­ous­ly per­ceived cakes. (Image cour­tesy: Smith and Koss­lyn, Cog­ni­tive Psy­chol­o­gy: Mind and Brain).

Cat­e­go­riza­tion binds objects with sim­i­lar fea­tures or attrib­ut­es togeth­er into a gen­er­al con­cept and there­fore makes infer­ence pos­si­ble. Because cat­e­go­ry knowl­edge con­tains diverse infor­ma­tion that goes much beyond what is imme­di­ate­ly per­ceiv­able by the eyes and the sens­es, it enables one to draw use­ful infer­ences about things, peo­ple, and sit­u­a­tions in life.


Final­ly, the third valid means of knowl­edge is described as aga­ma: a mat­ter direct­ly per­ceived or inferred by an accom­plished per­son and taught in the form of words to trans­fer his knowl­edge into the lis­ten­er. The result­ing vrit­ti from that word, with its mat­ter and mean­ing as its con­tent, is the listener’s acqui­si­tion, aga­ma. These words from Vyasa’s com­men­tary lead to pro­found insights into the teach­ing-learn­ing process that would require the writ­ing space of an entire arti­cle. Let’s stop here for now!


Arya, Pan­dit Ushar­budh. Yoga Sutras Of-Patan­jali With The Expo­si­tion Of Vyasa. The Himalayan Inter­na­tion­al Insti­tute: Penn­syl­va­nia (1986).

Smith, Edward E., and Stephen M. Koss­lyn. Cog­ni­tive psy­chol­o­gy: Pear­son new inter­na­tion­al edi­tion: Mind and brain. Pear­son High­er Ed, 2013.

Hills, Peter J and Pake, Micheal J. Cog­ni­tive psy­chol­o­gy for dum­mies. John Wiley & Sons, 2016.

4 views0 comments

Recent Posts

See All
bottom of page