How should we treat yoniso manasi √kar?

I was wondering whether any of our Pali specialists had any thoughts on this.

In our source text, yoniso manasikāra in its various forms is sometimes treated as a compound, sometimes separately. There doesn’t seem to be any hard and fast rules behind this, although the verbal forms tend to be more resolved. It seems like a phrase that is moving towards becoming a term.

Sometimes it seems better to consider it as a single compound.

yonisomanasikāramūlakā dhammā


ayonisomanasikāraṁ pahātuṁ

But in the same sutta we have:

ayoniso manasikāraṁ pahātuṁ

Sometimes, especially with verbal forms, it’s fully resolved:

yoniso manasi karoti

Sometimes partly:

yoniso manasikaroto

Where yoniso is used with other terms it seems clear enough that it is separate. And that is sometimes clearly the case here too:

Ayoniso, bhikkhave, manasi karoto

yoniso ca manasi karoti

For the latter, we also find it partly resolved.

yoniso ca manasikāroti

Generally speaking, the usages in our texts seem sensible, if not entirely consistent. I kind of feel like we should be consistent. But I’m wondering if there is a clear grammatical basis for preferring one approach over another.


Is your intention to standardise this? I am not clear why you would not leave it as you find in the manuscripts. I mean, language is always evolving. As I see it, the inconsistencies we see here are just part of that evolution. Is it problematic if our texts reflect the evolution of the Pali?


Maybe. It affects things like search engines and the like. Generally our edition aims at standardization, which then becomes an assumption. It was, in fact, because one of our tools turned up one of these inconsistencies that I started looking at it.


OK, a few more cents worth! The above examples make it fairly clear that yoniso originally was not part of a compound. That’s because you have words like bhikkhave and ca intruding in the middle. We then need to explain the compounded forms.

It seems they have at least two different sources. The first is where they function as bahubbīhi compounds, that is, as adjectives qualifying some other word. This is what happens in your first example, yonisomanasikāramūlakā dhammā and presumably also with ayonisomanasikārahetu, which seems to qualify diṭṭhi. In these cases the compounded form is required. The second possibility is that compounding is a later development. This might be what we see in the case of ayonisomanasikāraṁ pahātuṁ.

With the second possibility and concerning the negative expression ayoniso, the compounded form prevails slightly over the uncompounded form, with roughly 43 cases of the former versus 27 of the latter. Perhaps this ambivalence has to do with a lack of clarity as to what actually is being negated with the initial a. It seems natural to see the whole expression as negated, whereas the influence of the positive form may have resulted in the breaking up of the compound. So perhaps we are seeing two countervailing tendencies. But I am just guessing really, because it is far from clear what is going on.

A more detailed analysis reveals the following. It seems that when ayoniso manasikāra is used “independently” (e.g. as part of a list) or in a verbal form, then the words are split up. But when ayonisomanasikāra functions as the patient of a verb, as in ayonisomanasikāraṁ pahātuṁ then in most cases it is compounded (although there is one rogue uncompounded case: ayoniso manasikāraṁ pahātuṁ). In other cases again ayonisomanasikāra is part of larger compound, in which case the compounded form is probably used for the sake of clarity, for instance with ayonisomanasikārasuttaṃ or ayonisomanasikārabahulīkāro. Then again, there are cases where the compounded form is used in independent constructions, just like the uncompounded form, especially in the Aṅguttara Ekanipāta.

So the overall the situation seems quite confused. I doubt you will be able to find any consistent patterns. Argh!!! (Where is that emoji for tearing your hair out? Such a useful emoji, especially for monks!)