"Data" versus "metadata" doesn't go to the definition of "reasonable." It goes to the definition of "their" in "their persons, houses, papers, and effects..."
The basic distinction is your data versus data that's about you. If I own a store, and make a note that 001sky came in to buy a box of cookies at 2:00pm on January 23, 2005, that's my paper. It's not your paper. I can claim 4th amendment protection if the FBI breaks into my store and takes that note, but you can't. That's the legally relevant distinction between data and metadata.
> "Data" versus "metadata" doesn't go to the definition of "reasonable." It goes to the definition of "their" in "their persons, houses, papers, and effects..."
Not really, its still the persons, houses, papers, or effects of "the people" no matter which of the people it belongs to.
There's this weird idea which seems to have emerged from the use of the exclusionary rule as the main (and pretty much exclusive) remedy for violations of the Fourth Amendment that the Fourth Amendment only applies when the person about whom information is sought is the person whose person, houses, papers, or effects are being searched or seized, but there is nothing in the text of any reasonable context of the Constitutional language itself to support that, its simply a byproduct of an inadequate remedy (which, in fact, that aspect reveals to be inadequate) that seems to have worked its way back into the way that the language is read.
First, the 4th amendment isn't worded as a restriction on the actions of the government, like many other "rights" in the Constitution, but as a personal right. While an individual can't consent to say the government's exercise of an unconstitutional power, an individual can waive a personal right. For example, if the government eminent domains my property to build a highway, its not a violation of the Constitution if I waive my right to just compensation. So while the call metadata is clearly protected by the 4th amendment as to AT&T, its their personal right and they can waive it.
Second, ("th[e] weird idea... that the Fourth Amendment only applies when the person about whom information is sought is the person whose person, houses, papers, or effects are being searched or seized") doesn't stem from the exclusionary rule, but rather the subpoena power. The subpoena duces tecum (subpoena for the production of documents) predates the 4th amendment by quite a bit, appearing in England in its modern form sometime in the 1600's. While an overly broad or abusive subpoena can be a 4th amendment violation (but against the person at whom the subpoena is directed, not against the subject of the documents sought), they are presumptively reasonable.
Thes boundaries are blurred quite easliy with sufficiently advanced technology. That is the point. I can FLIR your house from a public location. Or isps and email providers could store every message (full data) just like the metadata and now 'own' a free copy. Arguably gmail has walked this line in their argument that scanning the metadata to route the message and scanning the content for meaning are equivalent interrogations of bits. But there are strong arguments about piercing of protected speech or inpinging on the veil privacy in the home, using technology to blunt force your way in are limited by jurisprudence already. That is why the surveilance by artifacts of speech and surveilance of actual speech is not such a simple issue.
There are situations in which technology changes things and the law has adapted (e.g. using an infrared camera to look into a house from a public location is a 4th amendment violation), but this is not one of those situations. At least when it comes to call metadata, technology really has nothing to do with it. Whether AT&T records it on magnetic tape or quantum bits, it's still their records, generated by their equipment, kept for their purposes, that you never have access to. These characteristics aren't an artifact of the technology, they're reflective of the fundamental nature of the information in question: it's AT&T's information about you, not your information.
At least when it comes to call metadata, technology really has nothing to do with it
I'm not sure this is logically possible. The metadata retention by the telco's is an artifact of their business process (minimize fraudulent billing liability)[1]. A similar process did not exist for mail until the government mandated and paid for it. Now, every piece of mail is photgraphed ("to collect the 'metadata' ")and the images are recorded. That notwithstanding, the entire notion of such meta-data is strictly an artifact of technology. There is no analogue in pure verbal speech, in other words.
[1] There is no need for them to retain it or to otherwise "posess" it in an-other-than-temporary-like-snapchat-way.
The fact that metadata is generated and collected may be an issue of technology, but that doesn't change the essential characteristic of whether it's your data, or data about you. E.g., in 1776, there were no video cameras in public places. But technology doesn't change the essential character observing you in a public place versus in a private place.
> That notwithstanding, the entire notion of such meta-data is strictly an artifact of technology. There is no analogue in pure verbal speech, in other words.
Sure there is! Before mail was scanned for delivery purposes, you could always have called a postal worker to testify about what he read on the outside of the envelope. Before store cameras, you could have always called a store worker to testify about when he saw 001sky enter and leave the store and what he saw 001sky buy. Before automated call tracking, you could always have called a phone operator to testify about when 001sky placed a call. Technology makes it easier to collect metadata, but it does not turn "data" into "metadata."
You make a strong case for certain points here and I don't dismiss them. But I also think several points are blurred. That is perhaps inevitable and that is the point I am making. for example
(1) That doesn't change the essential characteristic of whether it's your data, or data about you. The datacard in my FLIR camera is (a) my data; and (b) about you. Neither of these qualifications are insurmountable. The courts have not tripped over this technicality, because the <purpose> of using the FLIR instrument is not considered a reasonable use of technology, when it has the impact of making surveilance <unavoidable> per-se.
(2) The post office example. This raises two issues: (a) technology; and (b) scale. Mail by post is distinct from verbal speech and generates metadata because of the need to address an envelope. Its low technology, but this is an artifact of the technology. With respect to scale, what is 'reasonably' to collect is a function of resource expenditure. As the cost of collection decreases, the amount of data 'reasonably' collected <increases>. But that is beside the point of where my <reasonable expectation of privacy> has changed or stayed the same.
(3) Scale, generally speaking. eg: Technology makes it easier to collect metadata, but it does not turn "data" into "metadata." The express purpose of the meta-data programs is to turn <meta> data into <data>. This is sort of the fundamental question at the heart of the matter, from a constitutional perspective. The purpose of these systems is to surveil beyond the scope of what would normally be perceived as reasonable. In that regard, it is aking to using FLIR to make surveilance <unavoidable per-se>. Or at least that is an argument which should be ajudicated. Perhaps by legislation, as well, once a more coherent framework has been set forth by the courts. In otherwords, now that these 'edge cases' are being swept up into the norm.
The basic distinction is your data versus data that's about you. If I own a store, and make a note that 001sky came in to buy a box of cookies at 2:00pm on January 23, 2005, that's my paper. It's not your paper. I can claim 4th amendment protection if the FBI breaks into my store and takes that note, but you can't. That's the legally relevant distinction between data and metadata.