This is the first entry in my attempt at an annotated translation of Gaṅgēśa’s Tattvacintāmaṇi (TCM). In this post, I will translate and discuss section I, chapter I of the first volume, Perception [Pratyakṣakhaṇḍaḥ]. In other words, I’m skipping the preamble, A Theory of Good Beginnings [Maṅgalavādaḥ], and starting with the Prior Position [Pūrvapakṣaḥ] section of the chapter Detecting Evidentiality [Prāmāṇyajñaptivādaḥ].
My source text is the 1973 Tirupati edition (pub. Kendriya Sanskrit Vidyapeetha, ed. N.S. Ramanuja Tatacharya). I am reading the text along with the Prakāśa and Nyāyaśikhāmaṇi commentaries; by Rucidattamiśra and Rāmakr̥ṣṇādhvarin, respectively. I also make use of Stephen Phillips’ english translation and commentary.
My aim is to create something like a self-study guide as a way for me to process what I’m reading as I work through this text. But, I also hope it will help anyone else wanting to study the TCM. Fair warning: these notes reflect my own limited understanding of the material and I can’t claim either completeness or accuracy.
So—aiming to raise up the Earth itself, mired in the murk of dispair, the all-compassionate sage invented Philosophy, most revered among the eighteen spheres of learning. With this in mind, and to stimulate thoughtful readers, he began his Formulae with the words: “the path to the ultimate good lies in understanding the sixteen principles, starting with the evidential instrumentspramāṇam”. And, because, among these principles, it is upon the instruments that all the rest depend; this is what we will now discuss.
But—it is not correct that the intellectual practice proceeding from an analysis of our evidential instruments is necessarily connected to the attainment of the highest good. For, it is not possible to analyse the instruments, in the first place. This, after all, depends on an analysis of evidentiality; and the latter—as we will demonstrate—cannot be accomplished either from an internalistsvataḥ or externalistparataḥ standpoint.
On the other hand—why even bother determining evidentiality? Actionpravr̥tti, as such, does not depend on this. All that matters, when embarking on some potentially expensive course of action, is beliefniścaya regarding the necessary factsartha. And, you cannot respond that when some P is such that doubtsaṁśaya that P or belief that not P is a negative condition on some action A, then belief that P is a reason for A, like entailment is for inferring. For, when there is doubt about evidentiality, it is because of the resulting doubt regarding the evidenced fact that there is no corresponding action. So, the former is not the negative condition, here.
No—because when we face doubt about our evidence; then, confirming that such and such really is the case, which is a precondition for unhesitating action, does depend on belief in evidentiality.
Then—let us take establishing this to be internal.
Here are our options:
The evidentiality of an observation
1. is detectedgrāhya by just that cognitive processsāmagrī which detects the observation and does not detect it as nonevidential
— Yes or No?
2. is detected by a cognition that does not derive from a cognition whose content is that observation
— Yes or No?
Because whereas, in an externalist account, a mental state must be observed prior to detecting its evidentiality being detected; in an internalist account, a cognition is detected as evidential, at the onset.
3. is detected by that which is just the detector of its own locusāśraya
— Yes or No?
4. is detected by that which is just the evidentiality’s own locus
— Yes or No?
— does this apply in every case?
However, because in cases such as observations stemming from mystical insight or our mental connection to universals, we find instances of mental states detecting their own evidentiality; the theses fail to fully differentiate themselves. Or else, you must qualify your assertion with the phrase “not caused by mystical insight, etc.”
No—because not everyone accepts that these things exist. And even if such things were accepted, since the claim is about detecting evidentiality, in general, and about whether evidentiality is detected by that which is just its own locus; it is not the case that theses fail to fully differentiate themselves.
So, then—is it or is it not the case that the evidentiality of the cognition ‘this is a pot’ is detected by that very mental state? Is it or is it not the case that it is detected by just its own detector?
Now, whether the evidentiality of a cognition is defined as the property of
1. not being modulated by a subject unlocated property, or
2. detecting an undetected fact, or
Assuming the universal factuality of observation, and given that observation is not a natural kind:
3. being nonrecollective awareness, or
4. being a structured mental state
it is not detected either by that cognition, itself, or by a perceptual state introspecting it. For, without awareness of the property "not presenting a subject nonpredicating property", etc.; one cannot be aware of its predication of a mental state.
Nor, under anti-perceptualism about introspection, can a cognition's evidentiality be detected via an inference from cognitivity; for the property 'being cognized', on its own, does not track with properties like "being evident". Nor does it work to modify the inference to 'being observed in a manner conducive to evidence'; since, such a property would not be manifest at the onset. Otherwise, even the detection of nonevidentiality would be internal, via an inference from 'not being observed in a manner conducive to evidence'.
Except—given some property P, evidentiality is defined as either
For, only from such a belief can there be confident action. This is the economical account and anything else introduces complications. And this, moreover, can be detected by the introspecting mental state. So, whether by an inner perception, an inference, or reflexive awareness; introspection detects as the content of a mental state both that an object is the locus of some property and that this object is presented under that property.
Mental states are determinednirūpyam by their content. So, the content of a first order mental state is present in the introspecting state. This being the case, when a mental state is experienced, in addition to an object T and a property P, the predication by P of T also features as its content. For, the latter is just as much a part of what defines the structure of that mental state.
Since the form of a relatum in awareness is delimitedavacchinna by the relation it partakes in, the introspecting state is determined by not just the relata, but the relation as well. In other words, the content of an awareness of, e.g., ‘this object in front of me’ appears in introspection as delimited by, e.g., ‘being silver’. Otherwise, introspection would take the form, ‘I am aware of this object in front of me and silver’. Not, ‘I am aware of this object in front of me as silver’.
If this were the case, then nonevidential states would also be detected as being evidential; since the introspecting state would have for its content predication by a property which is the content of an error. Nor could you respond that since evidentiality is not known at the onset, it would not be imposed on the introspecting state. After all, because the evidentiality of a cognition is detected by the mental state detecting that cognition; it is observed in introspection—like with awareness of an absence, where both the absence and the absentee are presented.
The nonevidentiality of a P-presenting cognition of a P-nonlocus is cognized externally. Since, so long as the error is not made explicit, that the object lacks that property is not the content of the introspecting cognition. Nor should you say that insofar as the content of introspection includes the full content of the introspected cognition, there would be a conflation of the cognition presenting that error with the error itself. After all, so long as no defeater appears, this is precisely what we expect. But, when we do encounter a defeater, the conflation is blocked; because, importation of the error from the introspected cognition ceases. This is also why there is no error when observing someone else’s mistake.
Truth is not just a matter of predicating one thing of another. Rather, it is a peculiar property of cognitive content which, though appearing in the first-order cognition, does not appear in introspection due the mental organ’s lacking independent access to nonmental facts.
No—because, insofar as it, too, is a part of the content of the first-order cognition; this “peculiar property of cognitive content” would be imported into the introspecting state, just like the subject and predicate. Moreover, we have reason to doubt the existence of such a property of cogntive content.
From this, our responses to the following objections are also clear:
Assuming introspection is not reflexive: the detection of evidentiality cannot be internal because a property cannot be detected without detecting its locus. Detecting on the basis of its structure that a mental state is auditory, olfactory, etc. does not mean there is awareness of it as evidential.
Assuming introspection is reflexive: insofar as a cognition only observes itself, it lacks the capacity to detect its properties. Or, if it does have this capacity; nonevidentiality would also be detected internally.
Assuming introspection is not reflexive: the introspecting cognition, whether inferential or perceptual, cannot detect evidentiality as exclusive of error, because of a lack of correlation. That is—if it was detected only where it was present, then the detection of nonevidentiality would also be internal. Moreover, neither inference nor inner perception are causally connected to such content.
So, evidentiality is defined either as
because this is what is detected by introspection, whether understood as reflexive awareness, inner perception, or inference. That is, because the introspecting mental state detects of a cognition that it is P-presenting and that it has T-content.
Moreover, if detection of evidentiality was external; then, with respect to the evidentiality detector; believing its content—including the relevent inferential sign, target, and entailment relation—would, in turn, depend on detecting it’s evidentiality. And, insofar as this too would proceed externally, it would trigger a regress in the direction of both process and result. If this were so, we could not detect evidentiality at all. As such, we can also prove by reductio that detection of evidentiality is internal.