Quote from Chicken_Champ >>

Nope, they attack once, using an Attack that allows them to attack once once or twice, due to their Extra Attack feature.

 If it weren’t for the “instead of” or “rather than” words in EA, you might have a point. Otherwise, I don’t see that your reading is required and is at least somewhat nonsensical.

once once

So, from reviewing this thread again it seems like there are two camps. Either you believe you can invoke an ability whenever you want or you invoke an ability only when it is actually used in practice. Each camp's methodology has different implications for how SM would work.

If the former, you can invoke extra attack any time you like or even continuously, the feature is just always on/active and thus you can attack twice when you use the Attack Action. This means when you use that Attack Action you now are simply eligible for two attacks, and can use one, two, or even none of them.

If you believe the later, then you are only considered to be using then extra Attack feature when you do benefit from its listed benefit, when you attack and actually do attack twice. You are only using the feature when you're actively benefitting from what it allows.

Nobody has proposed that you can take the attack action and take no attacks (or other activity swapped for attacks). Unless that is specifically your position nobody is saying that.

Quote from Chicken_Champ >>

Nobody has proposed that you can take the attack action and take no attacks (or other activity swapped for attacks). Unless that is specifically your position nobody is saying that.

Aren't you arguing for that position?

Quote from Chicken_Champ>>

In plain English, when somebody tells you that you "can" have more than one of something, that is understood to mean that number of something or less, in most contexts. See also, you "can" Action Surge twice per day as a Fighter at 17th level, the feature has not removed your ability to only Action Surge once per day instead. A Druid "can use [Wild Shape] twice," which obviously means they can also use it just once. 

Here you argue that Plain English meaning of "can" to mean you "can" use it "up to" the listed value. A Druid "can" wild shape 2 times per short rest. But they needn't wild shape exactly two times. They "can" wild shape only once, or even none at all.

I'm not sure there's any examples in 5E to be found of something that uses "you can ___ X times" to mean "only X times, no less", other than when X is "once." If you want to get all technical and demand to see where 5E spells out that "can ___ X times" should always be read "can ___ [up to] X times", you've got me, I doubt it spells that out as a cannon of general interpretation, if you don't already bring it with you as an unspoken understanding of how plain English uses "can."

Here you reiterate this idea again. A value less than the listed value is valid. You can do X, Y times. You can still opt to do it y-1 times, or y-2 times, all the way down to 0 times. Is that not your position?

Attack by default tells you to "make" one attack. Taking it to not make one attack is not allowed.

With Extra Attack, you may make more than one attack. 

That's been my position for several pages now,  and I've pumped the breaks on this slippery slope "well aren't you saying..." several times now. Nobody is (reasonably) arguing that you can take an action to do nothing. 

Quote from Chicken_Champ >>

 

That's been my position for several pages now,  and I've pumped the breaks on this slippery slope "well aren't you saying..." several times now. Nobody is (reasonably) arguing that you can take an action to do nothing. 

You ready an attack to shoot a guy with your longbow if he takes an attack action against your ally. Your potential target has Extra Attack. He is going to trigger your readied action, and fully intends to take out your ally with every attack he can muster. 

When, exactly, does your readied attack go off?

• After he takes the action, but before the attacks.
• After the first attack.
• After both attacks and the conclusion of his attack action.

You Ready needs to be triggered by something perceptible. If your DM rules that "is about to attack" is a sufficient Trigger, good for you, but I've more often seen "if they attack" ruled to be a more reasonable trigger since you can't necessarily see someone's intentions. On top of that, there is no attack action, up until the moment where they "make" at least one of the attacks, as I've been over. Your DM may think that "making" an attack is granular to the point where it's interruptable before the damage roll, before the attack roll, or heck even before a target is declared? But I would't bet on it. 

You take Attack when you meaningfully "make" one or more attacks (or attack replacements), not when you just declare your intentions. While there are indeed two camps about how flexible that "or more" is in relation to the Extra Attack feature, if you're describing a third camp where you can take Attack without or before making any attacks at all, it sounds to me like you're on your own in a third camp.

Quote from Chicken_Champ>>

You Ready needs to be triggered by something perceptible. If your DM rules that "is about to attack" is a sufficient Trigger, good for you, but I've more often seen "if they attack" ruled to be a more reasonable trigger since you can't necessarily see someone's intentions.

Answering the question would have demonstrated your position more expediently but sure, I see your point here.

On top of that, there is no attack action, up until the moment where they "make" at least one of the attacks, as I've been over.

Source? Sort of the crux of the matter. 

Your DM may think that "making" an attack is granular to the point where it's interruptable before the damage roll, before the attack roll, or heck even before a target is declared? But I would't bet on it. 

"DMs do the darndest things" TM.

You take Attack when you meaningfully "make" one or more attacks (or attack replacements), not when you just declare your intentions.

Not when you declare intention, no. You take it when you say you take it. That isn't a declaration of intent, but a declaration of action.

• "I intend to take the attack action" is nonbinding, you can change your mind, you're simply sharing your goal orientation with people at this point.
• "I take the attack action" is binding. You've taken the attack action now. This allows you to attack a target (more than once if you have Extra Attack).

While there are indeed two camps about how flexible that "or more" is in relation to the Extra Attack feature, if you're describing a third camp where you can take Attack without making any attacks, it sounds to me like you're on your own in a third camp.

Wouldn't be the first time, won't be the last. Popularity of the position aside, the game text is silent on the exact timing.

Taking the attack action and making an attack are two distinct concepts in 5e. They can be interrelated, but not always are. 

• We've already demonstrated you can take the attack action and then never actually "make an attack" by substitution. Yes? Yes.
• We've covered ways to make an attack that have nothing to do with taking the Attack Action. Yes"? Yes.
• So there is a distinction between them, they are discreet.

So, what is the timing? Which comes first? Do you take the Attack Action before you get to Make the Attack? Do you Make the Attack before you have taken the Attack Action? Which comes first?

• Since making the Attack is granted by the act of taking the Attack Action, taking the Attack Action must necessarily come first.

Meanings for "Take" and "make", that's going to be a plain English question. What does it mean to "take" the Attack action, if not something resembling "doing the things it lets you do?" I'm not sure what you'd suggest instead, but it sounds like you're angling for "Saying 'I take the Attack action'  is taking the Attack action, even if you don't make progress towards making any attacks" which would be a not particularly likely to impress the DM definition, if you ask me. 

Quote from Chicken_Champ >>

Meanings for "Take" and "make", that's going to be a plain English question. What does it mean to "take" the Attack action, if not something resembling "doing the things it lets you do?" I'm not sure what you'd suggest instead, but it sounds like you're angling for "Saying 'I take the Attack action'  is taking the Attack action, even if you don't make progress towards making any attacks" which would be a not particularly likely to impress the DM definition, if you ask me. 

The rules aren't entirely silent on this, I am surprised and pleased to report.

You can simply decide to take your action before your movement - this is explicitly a matter of simply deciding to do so.

So, that establishes you can simply decide (i.e. declare, since you have to inform your DM what your decision is for the game to work) to take your action.  Nothing else is required.  Determining what action you are taking may well be weedy, but taking an action is as simple as deciding you're going to.

The Attack action states "you make one melee or ranged attack" - note the lack of the word "can".

Now, that might imply you decide to act, do the thing, and the DM works out what action you took from the thing you did - but Bladesingers make that impossible, because a Bladesinger that casts a cantrip may well want to attack afterward, which they can only do if this is an attack action, or may in general have some ability they picked up somewhere requiring this to be a Cast a Spell action.  Point is, the game breaks down if that's how it works, so it can't work that way. 

Another possibility is that you decide to act and then declare an action to be taking, which you are then committed to.  For example, if you pick the Cast a Spell action, you are committed to casting a spell.  Same thing for Attack.  Etc.  Here, the game does not break down - e.g. Bladesingers can pick Attack, then later, when making the attack they're committed to, their subclass feature will let them swap to a spell and nothing breaks down.

I personally can't think of a third possibility, but there may well be one.  Barring some other possibility, that would suggest it must be option 2 above, since it can't be option 1 for reasons stated.

So since it's explicit that actions are declaration based - you take an action by saying you're taking an action - and my interpretation holds that typing the action is also declaration based, as soon as you make the declaration, you are taking the Attack action.  That should qualify for Shield Master's shove.

If someone has a competing interpretation, I am super open to it - it's deeply invalid to try and prove anything based on "I can't think of anything else".

Quote from quindraco>>

The rules aren't entirely silent on this, I am surprised and pleased to report.

You can simply decide to take your action before your movement - this is explicitly a matter of simply deciding to do so.

So, that establishes you can simply decide (i.e. declare, since you have to inform your DM what your decision is for the game to work) to take your action.  Nothing else is required.  Determining what action you are taking may well be weedy, but taking an action is as simple as deciding you're going to.

On your turn, you can move a distance up to your speed and take one action. You decide whether to move first or take your action first. Your speed--sometimes called your walking speed--is noted on your character sheet.

The most common actions you can take are described in the Actions in Combat section. Many class features and other abilities provide additional options for your action.

The Movement and Position section gives the rules for your move.

You can forgo moving, taking an action, or doing anything at all on your turn. If you can't decide what to do on your turn, consider taking the Dodge or Ready action, as described in "Actions in Combat."

The Attack action states "you make one melee or ranged attack" - note the lack of the word "can".

Now, that might imply you decide to act, do the thing, and the DM works out what action you took from the thing you did - but Bladesingers make that impossible, because a Bladesinger that casts a cantrip may well want to attack afterward, which they can only do if this is an attack action, or may in general have some ability they picked up somewhere requiring this to be a Cast a Spell action.  Point is, the game breaks down if that's how it works, so it can't work that way. 

Another possibility is that you decide to act and then declare an action to be taking, which you are then committed to.  For example, if you pick the Cast a Spell action, you are committed to casting a spell.  Same thing for Attack.  Etc.  Here, the game does not break down - e.g. Bladesingers can pick Attack, then later, when making the attack they're committed to, their subclass feature will let them swap to a spell and nothing breaks down.

I personally can't think of a third possibility, but there may well be one.  Barring some other possibility, that would suggest it must be option 2 above, since it can't be option 1 for reasons stated.

I mean, at least you start off here by acknowledging what the feature actually says (that taking the Attack action involves "doing the thing" (i.e., "make one melee or ranged attack"), but I don't see how you then transition to that being only one possibility. With the Attack action, you make one attack. Other features might modify that! Like Extra Attack permitting (not requiring) you to make more than one attack, or a manouver being made in place of an attack. But what is the feature that permits you to make less than one attack/attack replacement? There is no such feature, and it need not be considered as a meaningful "possibility" as a result. There is no conflict with the Bladesinger feature in requiring you to make one attack cast a cantrip in place of an attack in order to have taken the Attack action. It just... I am very unclear and getting increasingly frustrated as to what straight line folks think is drawn between that and making an Attack-action-triggered-shove before an Attack action has been taken. WTF. 

So since it's explicit that actions are declaration based - you take an action by saying you're taking an action - and my interpretation holds that typing the action is also declaration based, as soon as you make the declaration, you are taking the Attack action.  That should qualify for Shield Master's shove.

If someone has a competing interpretation, I am super open to it - it's deeply invalid to try and prove anything based on "I can't think of anything else".

The competing interpretation has been laid out ad nauseum for seven pages now, as well as being the obvious and intuitive meaning of "take" that everyone should wake up in the morning with before deciding to get up to any shenanigans, and this tortuous justification for "you can take an action without doing anything yet" has zero textual support. 

I'm with Chicken_Champ there's no declaration step, you take the action and resolve it altogheter. So when you take the Attack action, you make an attack. It's the very definition of it.

Actions in Combat: When you take your action on your turn, you can take one of the actions presented here, an action you gained from your class or a special feature, or an action that you improvise. Many monsters have action options of their own in their stat blocks. When you describe an action not detailed elsewhere in the rules, the DM tells you whether that action is possible and what kind of roll you need to make, if any, to determine success or failure.

The problem is really with attack, because they decided to make it non-atomic. There aren't other actions that sensibly allow you to do something else in the middle of the action.

Quote from quindraco >>

Now, that might imply you decide to act, do the thing, and the DM works out what action you took from the thing you did - but Bladesingers make that impossible, because a Bladesinger that casts a cantrip may well want to attack afterward, which they can only do if this is an attack action, or may in general have some ability they picked up somewhere requiring this to be a Cast a Spell action.  Point is, the game breaks down if that's how it works, so it can't work that way.

Another possibility is that you decide to act and then declare an action to be taking, which you are then committed to.  For example, if you pick the Cast a Spell action, you are committed to casting a spell.  Same thing for Attack.  Etc.  Here, the game does not break down - e.g. Bladesingers can pick Attack, then later, when making the attack they're committed to, their subclass feature will let them swap to a spell and nothing breaks down.

I personally can't think of a third possibility, but there may well be one.  Barring some other possibility, that would suggest it must be option 2 above, since it can't be option 1 for reasons stated.

 I'd say that your option 2 make things break down too since it allows you to count as doing something without ever doing it. And in the case of SM that is explicitly counter to the change in the ruling, according to Crawford and the SAC at least.
Frankly I don't see why you would think this is a reasonable option to begin with. If you say you do "A" and the immediately start to do "B", why would that mean that you have done "A". Especially when you want to be able to come back later and actually do "A" then. I might be wrong here but it seems to me that the only reason why one would want to do it that way is to get around the "If you do A, then you can do B" restriction on features like SM.

IMO there is a sensible and workable middle way between your options, that you act as you like and, if extra clarity is needed, declare what action you are taking or what feature/trait you are using or whatever special rule allows you to do what you are doing. So basically your option 1 just with a little bit of cooperation between player and DM to minimize misunderstandings.

Quote from Ravnodaus >>

If the former, you can invoke extra attack any time you like or even continuously, the feature is just always on/active and thus you can attack twice when you use the Attack Action. This means when you use that Attack Action you now are simply eligible for two attacks, and can use one, two, or even none of them.

If you believe the later, then you are only considered to be using then extra Attack feature when you do benefit from its listed benefit, when you attack and actually do attack twice. You are only using the feature when you're actively benefitting from what it allows.

Yea I would fall in the former camp here, that seems to me to be the intuitive way. And there are more features than Fighter EA or BSEA where this matters.

Say that I'm playing a Monk, do I need to activate my "Unarmored Movement" feature to be able to use the extra speed? Is my speed the standard 30 (or whatever depending upon race) and I only get the extra 10-30 if/when I need it?  If so how does that affect any dash actions I take? And do I need to use some of that extra unarmored speed to be able to use the "move along vertical/liquids part of the feature?

Also I have to agree with C_C on the "do nothing attack action" thing. You can certainly take the attack action and do things that aren't attacks if you have the features that allows for such swapping (Bladesinger or Battle master or such) but you cannot take the attack action and then not do anything at all.

Quote from Chicken_Champ >>

It just... I am very unclear and getting increasingly frustrated as to what straight line folks think is drawn between that and making an Attack-action-triggered-shove before an Attack action has been taken. WTF.

Everything happens in an order. Things in 5e can't happen simultaneously. And, in fact, if things were to trigger simultaneously the order in which they resolve is decided by whoever's turn it is.

• Making an Attack. This is a thing.
• Taking the Attack Action. This is a thing.

Which happens first? When you take the attack action to make an attack... which happened first? Did you take the action first, or did you make the attack first? Sequentially, one happened before the other.

You seem to have settled on the ruling that the 'Taking the attack Action' happens after the "Making an Attack'. Is that correct?

If that is your ruling, then you'd be correct in saying you must make the BA Shove after the first attack is made from the attack action.

But. If someone was to rule the opposite, that "Taking the Attack action" comes first, is the prerequisite for "Making an Attack" then, well, the opposite is true. You could take the attack action, making you now eligible for the BA Shove, and follow that shove up with the "Making an Attack" immediately after it.

The rules aren't entirely clear which comes first. In my mind, it makes sense that you must take the action before you can make the attack. Taking the action is what allows the attack. So it seems clearly sequential. You argue otherwise, and while I disagree with your conclusion your argument isn't logically flawed in any way. It is a good argument.

The biggest issue is the game doesn't explicitly tell us which comes first. And so, we can either come up with non-text based rulings to justify which we think comes first, or we can default to the method the game says we should determine simultaneous events, that leaves it up to whoever's turn it is to decide.

Quote from Ravnodaus>>

When you take the attack action to make an attack... which happened first?...Sequentially, one happened before the other...The rules aren't entirely clear which comes first. 

The rules only seem unclear because you're assuming they're two sequential processes. But actions are things you do, so the action and its effects are one and the same. You've taken the Attack action if and only if the attack has been made, because that's the action's definition. There's no order there.

It's only Extra Attack that creates ambiguities by letting an action be in an unfinished state, because now you have to decide if "taking the Attack action" means having started it, or having executed it to completion.

Quote from InquisitiveCoder >>

Quote from Ravnodaus>>

When you take the attack action to make an attack... which happened first?...Sequentially, one happened before the other...The rules aren't entirely clear which comes first. 

The rules only seem unclear because you're assuming they're two sequential processes. But actions are things you do, so the action and its effects are one and the same. You've taken the Attack action if and only if the attack has been made, because that's the action's definition. There's no order there.

It's only Extra Attack that creates ambiguities by letting an action be in an unfinished state, because now you have to decide if "taking the Attack action" means having started it, or having executed it to completion.

 

Extra Attack more than creates ambiguity. It proves that interpretation is wrong.

The problem is really 'when is the condition filled'.

Attack means that, when you do it, you perform an attack (or something that another rule allows you to use as a replacement for an attack). There is no 'may' in there. Thus, attack and Making an Attack are not actually separate things; your action is 'I attack the orc with my sword'.

The problem is that a bonus action, unless the action specifies otherwise, may be taken at any time during your turn, and thus it can interrupt your attack. If Shield Master didn't have a prerequisite, it would be totally legal to say "I attack the orc with my sword, but before actually attacking, I try to knock it prone with my shield". So the question is, when does the precondition "If you take the Attack action on your turn" actually get fulfilled.

Quote from Pantagruel666 >>

The problem is really 'when is the condition filled'.

Reading only RAW without any other outside influence, a particular reading of the feat's wording could be considered ambiguous, which is why JC weighed in originally (and then again later reversing himself, lol). The feat says "if" rather than "when" or "if/then" and "if" by itself just indicates dependence rather than order of operations. Sure, this is debatable, but you could take the attack action on your turn right after your bonus action, and you have technically satisfied the "if" portion of the statement by the time your turn ends. And before JC's tweets, this was the general use of the feat because it's mechanically the better option.

JC says, "No general rule allows you to insert a bonus action between attacks in a single action. You can interrupt a multiple-attack action with a bonus action/reaction only if the trigger of the bonus action/reaction is an attack, rather than the action." If you fall back on the wording of the bonus action itself, it says "You choose when to take a bonus action during your turn, unless the bonus action's timing is specified", so it's fair to take exception with his statement. I like looking into the intention behind the game designer's choices, so I respect where JC says, "if a feature says you can do X as a bonus action if you do Y, you must do Y before you can do X. For Shield Master, that means the bonus action must come after the Attack action. You decide when it happens afterward that turn." That lets you know that the bonus action shove is intended to occur after the attack action, and it explains why. That statement really should be formed into an errata or a SAC entry though because it's fairly encompassing in its scope.

Ultimately, I think it's fair to just ask yourself whether the attack action is finished or whether there are still attacks coming and use that as your guideline. Or just lean into the RAW ambiguity and say that none of that stuff is in the rules, and it's a LOT more fun to knock someone down with your shield and then beat on them while they are on the ground. And you're right. It is more fun.

It would be quite easy to errata the feat to say: 

• AFTER you take the Attack action on your turn, you can use a bonus action to try to shove a creature within 5 feet of you with your shield.