This is for a fairly absurd situation in which a party member (the cleric) has died and neither of the others can resurrect him, but one of the surviving party members is a wizard that can cast simulacrum.

Another player's idea was to cast simulacrum to create a simulacrum of the cleric (with "half the creature's hit point maximum" - which seems independent of its current HP) and then have the simulacrum resurrect the cleric.

The description of simulacrum says:

You shape an illusory duplicate of one beast or humanoid that is within range for the entire casting time of the spell. [...] It appears to be the same as the original, but it has half the creature's hit point maximum and is formed without any equipment. Otherwise, the illusion uses all the statistics of the creature it duplicates.

Ultimately, it seems to be a question of whether creatures retain their type after they die (i.e. when they are a corpse).

A few of Crawford's tweets have stated that a corpse is an object that is no longer a creature, suggesting that spells that target creatures would no longer work on corpses...

...but then spells like revivify explicitly describe the target as "a creature":

You touch a creature that has died within the last minute. That creature returns to life with 1 hit point. This spell can't return to life a creature that has died of old age, nor can it restore any missing body parts.

And spells like animate dead even suggest that creature type is retained:

Choose a pile of bones or a corpse of a Medium or Small humanoid within range.

As a result, my findings so far seem somewhat inconclusive and contradictory. The intent obviously seems to be that simulacrum is not intended to create a duplicate of a corpse (which seems pretty ridiculous), but I'm not sure whether it'd work at least by RAW.

By the rules as written, could simulacrum be used to create a living duplicate of a dead creature (a corpse)?


3 Answers 3


No, as your corpse is not you

The problem is not with the creature type restriction, but with the requirement that:

one beast or humanoid that is within range for the entire casting time of the spell

The creature you copy has to be present. So then, is the corpse of a creature still itself?

Sadly both the wording of the books and the "errata tweets" tend to be rather sloppy, and you have found such a case. So a DM has to make a ruling.

It would unnecessarily complicate things to rule that your corpse is still you. It raises questions like: Then is your soul, now departed, also you? When does your body stop being you? When it is just a skeleton? When incinerated?

Ruling that the corpse of a creature is now an object and that the reviving spells should have been worded "you touch the corpse of a creature that has died" poses much fewer problems, if any. I strongly recommend this ruling.

  • \$\begingroup\$ Assuming there can be only one you, True Resurrection proves your point: there can be a corpse, and a living version, without any connection other than the 'youness', no part of the dead one was used in creating the living one, therefore no transfer of identity occurs. \$\endgroup\$
    – Martin S.
    Oct 8, 2018 at 10:24

The only way there's a contradiction in the RAW is if "the corpse of a humanoid" implies that its subject is still a humanoid, or "a creature that died" implies that its subject is still a creature. Both are tenuous interpretations that fall apart if you try to apply them consistently.

Consider a longsword. An weapon with the sub-category 'longsword' has unambiguous properties per the rules: it's a versatile, martial melee weapon that does 1d8 or 1d10 slashing damage. Are "the pieces of a broken longsword" a longsword? Is "a longsword that was broken" a longsword? (In this case, it's potentially even still a weapon, albeit an improvised one.)

The change, from whole to broken, from living to dead, is transformative. It moves the object into a new category, the properties of which may be parallel to, but do not import any rules-meaning from its previous category.

  • \$\begingroup\$ This makes a fairly compelling point, though I disagree with part of it. I agree that "the pieces of a broken longsword" would not be a longsword - but describing something as "a longsword that was broken" would indeed suggest that it is currently classified as a longsword. (I assume the syntax should be "a longsword that broke" to match the syntax of the example in my post, though it wouldn't change the point I'm making in my previous sentence in this comment.) \$\endgroup\$
    – V2Blast
    Oct 7, 2018 at 18:22
  • 2
    \$\begingroup\$ Mandatory allusion to the shards of Narsil in 3, 2, 1 ... \$\endgroup\$ Oct 8, 2018 at 12:10
  • \$\begingroup\$ It is not that it would not be classified as a longsword, it's that longsword as a subcategory of broken things does not automatically inherit the properties of longsword as a subcategory of whole things. In the D&D rules, it's safe to assume that Humanoid, used alone, means "creature of type humanoid", but when the usage is qualified, the creature part can no longer be implied. \$\endgroup\$ Oct 9, 2018 at 22:35

I think the simulacrum spell would not work in this case.

Animate dead suggests the corpse needs to be the corpse of a Small or Medium humanoid, not that the corpse itself is one.

I think that revivify and other resurrection spells specify "a creature that has died" because a corpse technically is a creature that has died, but is still not a creature itself.

  • \$\begingroup\$ Welcome to RPG.SE! Take the tour if you haven't already, and check out the help center for more guidance. \$\endgroup\$
    – V2Blast
    Dec 27, 2020 at 0:15

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .