Sage Advice Compendium Update 1/30/2019

[DND_Reborn]

While I think it would be great to have precise text in the PHB to support either position, IME so far the text is simply not there. If it was, people would have been using it by now.

Even if someone finds text they believe supports their position well, 9 times out of 10 the wording is likely ambiguous enough to still allow for different interpretations. Often this was done purposefully to allow the tables to run as they wanted to make the game the most fun for them, other times it was honest mistakes or oversights which are addressed in errata.

Resorting to other sources such as SA is fine, but even that is taken with a grain of salt by many and official only in the loosest of terms.

It is ironic that a couple days ago when we began our last session and a player was making his Paladin/Rogue, I brought up the text of Shield Master and of the other two players (I'd already discussed it at length with our DM), one felt the shove had to come at the end, the other believed you could do it at the beginning if you wanted to (but then you had to use the Attack action afterwards, no other actions were permissible).

Isn't that funny? Two uninformed opinions in our group and one chooses the attack then shove and the other the shove when you want but must attack at some point?

When I told them about this thread and we had our discussion, we compromised on the attack, shove, and maybe more attacks if you have them model.

Sorry about wandering a bit, but my point is that since the text we have available isn't definitive unless you choose to follow SA and JC's ruling, this thread will never end...

Maybe that is the point, though, huh?

 

[Arial Black]

This is an interesting argument.

Huzzah! At last!

You say there's no causality in "If X, Y" but then say that if Y occurs, it causes X to occur simultaneously, if X has not already occurred on its own. Interesting, but not consistent with itself.

I agree! That's because I only hold one of these positions (the correct one!), that "if...then.." statements are not statements of causality. I mentioned this in a previous post, citing Wikipedia's essay on Causality and contrasting with conditionals. This means that the 'effect' can come before the 'cause', depending on what the condition actually is, and since the conditional here is , "on your turn", then the "if...then..." form does not require the shield shove to come after the attack; the only requirement is that both take place on your turn.

However, I anticipated that some posters would continue to assert, erroneously, that this particular "if...then..." statement should be adjudicated as if it were a statement of causality. Boy, was my anticipation correct!

What I then showed was that, even with that interpretation, it still results in being able to resolve the shield shove before you resolve the attack.

So, either way, the shield shove can go before the first attack.

You are correct that "If X, Y" does not strictly imply causality, but you've gone further than that to claim it cannot be a casaul relatioshop, and this is wrong. Especially in this case. For "If X, Y" to be non-causal, X and Y must be independent of each other, or, at worst, both caused by some other Z. This is observed in that "If X, Y" does not imply Y cannot exist without X, ie, you can have a Y without an X, but if you have X, you will also have Y.

But, that's not the case with Shield Master because you cannot ever have the bonus action shove without the condition being true. This extra requirement is part of the general bonus action rule -- you don't have one unless something gives it to you -- which is a causal statement. This must be considered with Shield Master.

I'm not implying that it cannot be a causal relationship. I'm saying that because of the "if...then..." form, the relationship is not bound by Time's Arrow, while straight causal relationships must be. And in this case, the condition itself, freed from having to occur in a set order, instead merely needs to obey the condition! And the condition is merely that they both take place "on your turn".

Further, while duration of actions is meaningless within the 5e action structure (actions have no assigned duration), the turn structure does have a clear order. Therefore, since there is an order, and since actions a discrete elements (you pick from a menu of actions with unique results and can't mix and match results), and since the Shield Master bonus action's very existence is predicated on taking the Attack action, you must take the Attack action before you can use the bonus action shove.

No!

First, 'no' from the perspective that "if...then..." statements are not bound it time, just bound by the particular condition.

Second, 'no' from a perspective which tries to claim that this particular "if...then..." statement should be treated as if it were a straight statement of causality, if that claim were true then it must obey the requirements for a temporal order, and I know you agree that it should obey that temporal order!

But the crucial part is this: "statements of causality require the antecedent to precede or coincide with the consequent in time".

So even if you think that the 'effect' (the bonus action shield shove) cannot occur before the 'cause' (the attack), the two certainly can coincide!

Even in this interpretation, the very same moment you 'take the Attack action on your turn' you have simultaneously generated that bonus action shield shove. And because of the actual written rule stating that you can take your bonus action whenever you want during your turn as long as you have one to take, and you get this one at the very moment you 'take the Attack action', then, sure, under this interpretation you cannot take the bonus action before you get it, you certainly can take it at the very moment you get it! Because, "statements of causality require the antecedent to precede or coincide with the consequent in time".

Although you are free to choose to take the bonus action after the attack, and (under this interpretation) you cannot take the bonus action before the attack which 'caused' it, you are certainly free to take that bonus action at the same time as the attack which 'caused' it!

This also addresses the timing question. Since the bonus action shove is caused by taking the Attack action and since combat turns are ordered sequences, the Shield Master conditional is therefore a statement of timing.

Not quite.

It's not that "combat turns are ordered sequences", it's that the resolution of events is 'ordered', meaning 'resolved' one after the other.

So the two actions are certainly allowed to happen simultaneously. Actions like Dash, Dodge and Disengage could not function as we use them if two actions could not occur simultaneously!

However, things must be resolved sequentially!

In 5e, when things happen simultaneously on a creature's turn, they must be resolved sequentially even if they occur at the same time.

But who decides the order in which simultaneous things get resolved?

That's right! The player of the acting creature gets to decide the order in which they are resolved!

Here, you 'take the Attack action on your turn', which 'causes' the bonus action. That bonus action is generated at the same time as you take the Attack action.

Because 'cause' and 'effect' can coincide, and the rules say you can take your bonus action when you want, you can 'take that bonus action' at the same time you get it, which is at the same time as you 'take the Attack action'.

Since you chose to have the two things occur at the same time, you get to choose the order in which they are resolved. Therefore, you can choose to resolve the shield bash first, even though both the shield bash and the attack coincide.

[Edited to fix some spelling errors, but probably not all soellibg errors.]

 

[Asgorath]

...

Because 'cause' and 'effect' can coincide, and the rules say you can take your bonus action when you want, you can 'take that bonus action' at the same time you get it, which is at the same time as you 'take the Attack action'.

Since you chose to have the two things occur at the same time, you get to choose the order in which they are resolved. Therefore, you can choose to resolve the shield bash first, even though both the shield bash and the attack coincide.

Again, can you please show me the words in the Attack action definition that explain where the attacks happen after or separately from the action itself? Or the rule that says two things can happen at the exact same time? Or that performing an action or bonus action is independent of the resolution of that event? It's a turn-based game and all the language I've seen points to a strict ordered sequence of events that get resolved one at a time.

Example: You Shield Master shove first, an enemy uses their reaction to incapacitate you. You never made any attacks and thus never took the Attack action. How did you have the Shield Master bonus action on your menu of available options? If there's text in the PHB that says the attacks are separate from the action, or anything else you said, please quote it here.

 

[Arial Black]

What part of the wording of the Attack action makes you believe that the action is separate from the attacks it grants?

I'm not suggesting that the ability to execute those attacks is totally divorced from the Attack action which granted them, I'm suggesting that there is a difference between 'taking an action' and 'executing the stuff granted by that action', countering the claim that 'taking an Action' IS THE SAME THING as 'the effect of that Action'.

The rules for Extra Attack and the rules for Moving Between Attacks make it clear that you certainly can use your first attack to stab the goblin in the kitchen, then move into the dining room, then slash at the orc.

This means that, even though we don't know how many seconds go by between attacks, we do know that these attacks cannot have been simultaneous with each other.

IF we accept that 'taking an action' is the same thing as the stuff allowed by taking that action, that means that both of those attacks ARE the Attack action. Since those attacks were certainly not occurring simultaneously with each other, this means that the Attack action, with Extra Attack, is not necessarily a single, instantaneous event.

So, after the first attack on the goblin but before the attack on the orc, there is a gap in time into which one can 'take a bonus action', because there is a written rule which says you can take your bonus action when you like on your turn, and no written rule saying anything to the effect of, "...except during another action".

That says quite clearly that with the action, you make an attack. Extra Attack turns that one attack into multiple attacks, but at no point is there any mention of you being able to take the action first and the attacks later on in your turn.

The goblin and the orc say otherwise!

After you attack the goblin but before you attack the orc, have you met the condition, 'take the Attack action on your turn'? It's binary; it's either 'yes' or 'no'.

If it's, "yes you have met the condition", then you have met that condition before the second attack took place. This also means that the second attack cannot have been 'take the Attack action' itself, because if it was then you could not have satisfied the condition!

If it's, "no, you have not met the condition", then you did not 'take the Attack action', and are free to take whatever Action you want because you have an unspent action!

Which is it?

If you are tempted to wriggle on the hook by claiming that executing that first attack commits you to taking the Attack action, then by the same logic, taking the bonus action shield shove first also commits you to taking the Attack action! If you have no problem with the former, you can have no problem with the latter!

Surely if the attacks were separate, this action would include language like the Disengage action where it provides an effect for a duration, in both of these cases until the end of your turn? As it says above, the Attack action is the most common action taken in combat, so if it was supposed to work in the way you are describing, then I don't understand why the rules don't explicitly say that's how it works.

Either 'Actions are indivisible', or they are divisible! This applies to ALL Actions In Combat. It is the fallacy of Special Pleading to assert that all the OTHER Actions are "effects with a duration", but not the Attack action! There is no rules justification to treat them differently!

Either ALL Actions are 'effects with a duration' (so the Attack action allows you to execute your allowed attacks until you either run out of attacks or your turn ends, whichever comes first), i.e. 'actions are instantaneous declarations at the gaming table, with lasting effects', or ALL actions ARE their effects, and in that case if 'Actions are indivisible', then ALL actions are indivisible, rendering Dash, Dodge and Disengage all but unusable.

 

[Arial Black]

I'm not sure what you are trying to show here. This is a game where specific beats general, so yes, specific abilities and spells can override the general rules. There is no such time travel built into actions.

My point is that although there is no time travel in the game world going on with shield-bashing first, the fact that the game mechanic there seems to be time travel at the gaming table is completely fine, given that 5e has many such game mechanics that seem to work that way, and none of those involve time travel in the game world either.

It means that there is no valid 'time travel' objection to this game mechanic of shield-bashing first.

 

[Arial Black]

So again, specific beats general. The Attack action and your first attack are simultaneous per RAW. That's what "with" means. When you get a specific ability that attaches itself to the general rule via extra attack, it gets tacked on AFTER that first attack and then you get duration and divisibility, but only divisibility after the first attack.

I'm pleased that you agree that Actions are divisible. I am arguing against the 'Actions are indivisible' phantom rule which is being used to justify denying the shield shove between attacks.

I'm also arguing against other related things too.

 

[Arial Black]

There is nothing in RAW that even hints at the ability to declare a bonus action an attack action...

And I never wrote or thought such a thing.

...or that a bonus action somehow triggers the attack action.

Never thought that either. The Attack action 'causes' the bonus action, but because of the "if...then..." conditional, cause and effect are not fixed to a particular order, only to the condition 'on your turn'.

And even if you treat this trigger as if it were a statement of causality, you certainly can have cause and effect coincide, and choose to resolve the shield shove first.

So, either way, shield shove can come first.

 

[Maxperson]

My point is that although there is no time travel in the game world going on with shield-bashing first, the fact that the game mechanic there seems to be time travel at the gaming table is completely fine, given that 5e has many such game mechanics that seem to work that way, and none of those involve time travel in the game world either.

It means that there is no valid 'time travel' objection to this game mechanic of shield-bashing first.

The "time travel" comes in with the claim that if you use the bonus action first and then are prevented from ever taking the action, that somehow the action was taken before hand or during the bonus action anyway OR that the bonus action then becomes an action. The first is time travel, the second is Schrodinger's action, both of which are 100% unsupported in RAW.

 

[Maxperson]

I'm pleased that you agree that Actions are divisible. I am arguing against the 'Actions are indivisible' phantom rule which is being used to justify denying the shield shove between attacks.

I'm also arguing against other related things too.

Okay. I had already changed my position on the divisibility of action. They are divisible. What is not divisible is the the point between the start of the Attack action and the first attack. They come into being simultaneously. Once you add in the extra attack, the space in-between is open for things to happen.

 

[Maxperson]

Never thought that either. The Attack action 'causes' the bonus action, but because of the "if...then..." conditional, cause and effect are not fixed to a particular order, only to the condition 'on your turn'.

And even if you treat this trigger as if it were a statement of causality, you certainly can have cause and effect coincide, and choose to resolve the shield shove first.

So, either way, shield shove can come first.

Except that there is nothing to coincide with until after the first attack concludes. The Attack action does not begin until you hit step one of your first attack. Before step 1 of that first attack, no Attack action has even begun, let alone been taken. Once you have begun that first attack, you must resolve it before you do anything else, unless a special rule such as Sanctuary prevents you from doing so. Since no Attack action has even begun until the first attack starts and completes, there is no trigger for the bonus shove until that first attack completes.