2 5 Pāṇinian and Post-Pāṇinian Approaches to Derivations

In order to determine why post-Pāṇinian (both traditional and modern) scholars have misinterpreted Pāṇini’s rule 1.4.2, we need to understand that there is a fundamental difference between what I think are Pāṇinian1 and post-Pāṇinian conceptions of, or perspectives towards, the derivational procedure itself. I will explain exactly what I mean by this statement by means of examining six representative examples of SSRI from both Pāṇinian and post-Pāṇinian perspectives below.

39

Let us start with the latter. But before we examine these representative examples from the post Pāṇinian perspective, let me explain certain fundamental concepts which will help us understand this perspective better. Let us divide SSRI into two categories, namely ‘conflict’ and ‘non-conflict’.

In order to define conflict and non-conflict, we must first define blocking. Let us say that two rules X and Y are simultaneously applicable at step K2. We say that rule X blocks rule Y if Y will not be applicable at the following step (K+1) after the hypothetical application of X at the present step (K). Conflict is defined as an SSRI which involves some blocking. Non-conflict is defined as an SSRI which does not involve any blocking.

  • SSRI
    • blocking → conflict
    • no blocking → non-conflict

Note that, in my opinion, Pāṇini has not defined or discussed the categories ‘conflict’ and ‘non conflict’ in any way whatsoever, and he does not expect us to know about or use them either. Traditional scholars too have not made an explicit distinction between conflict and non conflict. In modern / western scholarship, the concept of ‘(rule) conflict’ has been widely used, but ‘non-conflict’ has not been used at all.

Then, the question arises: why have I made this distinction between conflict and non-conflict? I have done this to highlight that, for the most part, post-Pāṇinian scholarship has focused on conflict and has not paid much attention to non-conflict. Why is this the case? To answer this question, let us look closely at non-conflict, wherein the two rules X and Y do not block each other: if X applies at the present step, then Y is applicable at the following step, and if Y applies at the present step, then X is applicable at the following step.

Before we go further, note that ‘being applicable’ is different from ‘applying’. Consider the following situation:

[[40]]

Step 1: X → g, Y → h

Let us say Y applies at this step, changing h to h*. Now, at the following step, not only X but another rule Z too becomes applicable:

Step 2: X → g, Z → h*

Suppose that Z, and not X, applies at step 2.

Here, we see that, if Y applies at step 1, X is applicable at the step 2.3 However, X does not apply at the step 2. This is the difference between ‘being applicable’ and ‘applying’.

Now, let us go back to our conversation about why post-Pāṇinian scholarship does not take much interest in non-conflict. In most cases of non-conflict, if X applies at the present step, then Y is not only applicable but also applies at the following step. Similarly, if Y applies at the present step, then X is not only applicable but also applies at the following step. Thus, regardless of the order in which the two rules apply, one gets the correct form at the end of the derivation. This explains why the tradition can afford to overlook such examples of non conflict, which, as I said, constitute a huge majority of the set of all non-conflict examples.

However, there is a minority of examples of non-conflict wherein if Y applies at the present step, X is applicable at the following step, but does not end up applying at the following step. The tradition does take some interest in such examples of non-conflict, which constitute a very tiny minority of the set of all non-conflict examples.

[[41]]

Having defined both blocking and conflict, now let us look at how post-Pāṇinian scholarship views the following representative examples:

  1. (W, R→) k + l

If we apply R at this step, W will be applicable at the following step. R does not block W. If we apply W at this step, R will not be applicable at the following step. W blocks R.

We call this a case of asymmetrical or unidirectional blocking. Since this interaction involves blocking, this is a case of conflict. Such examples are of interest to post-Pāṇinian scholars.

  1. (S, V→) m + n

If we apply S at this step, V will not be applicable at the following step. S blocks V. If we apply V at this step, S will not be applicable at the following step. V blocks S.

We call this a case of symmetrical or mutual blocking. Since this interaction involves blocking, this is a case of conflict. Such examples are of interest to post-Pāṇinian scholars.

  1. (P→) e + (Y→) f

If we apply P at this step, Y will be applicable at the following step. P does not block Y. If we apply Y at this step, P will not be applicable at the following step. Y blocks P.

We call this a case of asymmetrical or unidirectional blocking. Since this interaction involves blocking, this is a case of conflict. Such examples are of interest to post-Pāṇinian scholars.

  1. (Q→) g + (X→) h

If we apply Q at this step, X will not be applicable at the following step. Q blocks X. If we apply X at this step, Q will not be applicable at the following step. X blocks Q.

We call this a case of symmetrical or mutual blocking. Since this interaction involves blocking, this is a case of conflict. Such examples are of interest to post-Pāṇinian scholars.

[[42]]

Post-Pāṇinian scholars are very interested in these four representative examples (REs). But one may ask: what about the remaining two REs? Let us look at them.

  1. (U, T→) i + j

If we apply T at this step, U will be applicable at the following step. T does not block U. If we apply U at this step, T will be applicable at the following step. U does not block T.

There is no blocking, so this is a case of non-conflict. The tradition does not think about or pay much heed to this kind of situation, for the most part.

  1. (O →) c + (Z →) d

If we apply O at this step, Z will be applicable at the following step. O does not block Z. If we apply Z at this step, O will be applicable at the following step. Z does not block O.

There is no blocking, so this is a case of non-conflict. The tradition does not think about or pay much heed to this kind of situation, for the most part.

Let us now summarize the relationship between blocking and conflict.

  • No blocking, Non-Conflict
  • Unidirectional blocking, Conflict
  • Mutual blocking, Conflict

[[43]]

Before we continue discussing these six examples from the post-Pāṇinian perspective, let us consider the Pāṇinian perspective on them:

  1. (W, R→) k + l

This is a case of SOI. Let us say W is more specific. Thus, W wins.

  1. (S, V→) m + n

This is a case of SOI. Let us say V is more specific. Thus, V wins.

  1. (P→) e + (Y→) f

This is a case of DOI. By 1.4.2, the RHS rule Y wins.

  1. (Q→) g + (X→) h

This is a case of DOI. By 1.4.2, the RHS rule X wins.

  1. (U, T→) i + j

This is a case of SOI. Let us say U is more specific. Thus, U wins.

  1. (O→) c + (Z→) d

This is a case of DOI. By 1.4.2, the RHS rule Z wins.

Note that in all six representative examples discussed here, Pāṇini does not require us to worry about what happens to the losing rule, for instance, P, in example 3: we need not be concerned about whether or not P is applicable at the following step, or whether or not P actually applies at the following step. In other words, Pāṇini does not use concepts like blocking and conflict to give instructions about dealing with SOI and DOI.

[[44]]

Even though Pāṇini does not use concepts like ‘conflict’ to give instructions about SSRI, and even though the tradition makes no explicit distinction between SOI and DOI, let us discuss both Pāṇinian and post-Pāṇinian concepts under one umbrella to understand this topic better. I have included both SOI and DOI examples because Pāṇini deals with them separately and have included examples of both conflict and non-conflict because the post-Pāṇinian approach subconsciously makes this distinction by focusing on conflict alone. Here is a summary of the examples:

RE4 Type Blocking Conflict
1 SOI unidirectional5 Yes
2 SOI mutual Yes
3 DOI unidirectional Yes
4 DOI mutual Yes
5 SOI none No
6 DOI none No

Representative examples 1, 2, 3 and 4 are of significant interest to post-Pāṇinian scholarship because they involve some kind of blocking, thereby constituting cases of conflict.

  1. When I make the distinction between Pāṇinian and post-Pāṇinian approaches in the following pages, it must be understood that by ‘Pāṇinian approach’, I mean ‘my interpretation of the Pāṇinian approach’. ↩︎

  2. In other words, let us say that there is an SSRI between X and Y. ↩︎

  3. So, going by the definition of blocking, Y does not block X. ↩︎

  4. RE = ‘Representative Example’. ↩︎

  5. Only a minority of cases of SOI involve unidirectional blocking. Most cases of SOI involve mutual blocking. ↩︎