Are receivers in a CAN network identified by their address?

ANSWER

Different CAN nodes must not transmit CAN data frames with the same identifier, but
the identifier of the CAN frames do not always contain any node ID for the identification.
As CAN is a message oriented network, the CAN identifiers are used to identify the message and it’s priority, not the network node. However, CAN based higher layer protocols such as CANopen or SAE J1939 implement a node-addressing respectively a combination of message prioritization and node-addressing to separate the different message types (network management, process and service data).

In CAN networks all messages are sent as a broadcast, and filters are only implemented on receiver-side. This means, every device has the chance to see all messages. It depends on the application of the device which data is used and which data is ignored. Devices use filters to only receive messages which they need. This keeps CPU load low.
Different CAN nodes must not transmit CAN data frames with the same identifier.