Chat, Texts, and Messages: Driving the Collapse of the Document in Discovery

This column looks at how technology is transforming evidence, litigation, and dispute resolution. This installment looks at how emails, PDFs, and other static files are no longer sufficient to tell the full story in modern litigation. Discovery is now a fragmented process spanning metadata, component parts, and other data sources. Past editions of the column can be found here.

For decades, litigation discovery was organized around the document. Find the email. Review the PDF. Produce the spreadsheet. Read the memo. Understand the file. That framework still matters. Static documents remain central to many disputes. But they are no longer always sufficient to tell the full story. We have moved beyond the four corners of the document—especially when the document no longer has four corners!

Modern business communication has become faster, smaller, and more fragmented. The information once contained in a single memorandum may now unfold across a collection of emails, texts, Teams messages, shared-document comments, calendar changes, phone calls, and system notifications. The communication has not disappeared. It has fragmented into many pieces in multiple forms.

That change has an important consequence for litigation: as the units of communication get smaller, the need for surrounding context gets larger.

From documents to fragments

The traditional discovery model assumed that the document was the natural container of meaning. A letter, memo, or email often contained enough information to understand who said what, when, and why.

Modern communications are different.

People increasingly communicate in short bursts. A statement is sent. Then clarified. Then corrected. Then answered. Then reacted to. The conversation may move from email to text, from text to a collaboration platform, and from there to a shared document or phone call.

This is not necessarily suspicious. It is just how people work now.

But it creates a discovery problem. A single message may be technically responsive and still be practically incomplete. Standing alone, it may not reveal the subject, the participants’ assumptions, the prior exchange, the document being discussed, or the action that followed.

The message may be the artifact—but the context may be the evidence.

I sometimes refer to this as the Mississippi River Delta problem. Upstream, the Mississippi River looks like a single channel. By the time that same water reaches the delta, it spreads into a network of smaller channels and distributaries. The water has not vanished. It has been redistributed.

Modern communication has undergone a similar shift. What used to be one larger unitized communication may now be distributed across many smaller pieces.

For litigation, the task is no longer simply finding the single channel. It is reconstructing the delta.

Why context becomes more important

This problem is anchored in information theory.

In the 1940s, polymath Claude Elwood Shannon’s work showed that information can be understood in terms of signals, uncertainty, and the transmission of symbols. Individual bits are small units. By themselves, they do not carry rich meaning. Meaning and efficiency come from structure around them.

Context helps make a message decodable. It reduces uncertainty. It allows the same symbol, word, message, or data point to convey more specific meaning.

Litigation has a similar problem.

A short text message saying “do it now” may be meaningless, misleading, or highly significant depending on the surrounding facts. What was “it”? Who sent the message? What happened immediately before? What happened immediately after? Was there a shared file, a phone call, a transaction, or a system event connected to that instruction?

The smaller the communication unit, the more likely it is that meaning depends on what surrounds it.

That is why modern discovery cannot always stop at the face of the document or message. In many cases, the surrounding system-generated data becomes essential to understanding the communication.

The rise of system-generated data

System-generated data is the information created by platforms, devices, and applications as people use them.

It may include timestamps, access logs, edit histories, version records, file paths, message IDs, attachment metadata, audit logs, login records, download activity, synchronization records, deletion records, and workflow events.

Often, a person does not write this data. The systems through which people communicate and work generate it. That makes it easy to overlook.

But in many modern disputes, system-generated data may provide the connective tissue between fragmented communications. It may show when a file was created, who accessed it, whether it was shared, whether it was edited, whether it was downloaded, what system activity followed, or how one artifact relates to another.

A PDF may show what a document looked like. Metadata may show when it was created. Version history may show how it changed. An access log may show who opened it. A chat message may show why it was sent. A transaction record may show what happened next.

No single item may tell the whole story. But together, they may make the story resolvable.

The practical tension: context vs. minimization

This creates a practical and legal tension.

Too little context may distort meaning. Producing a single text, chat message, or system notification may strip away the information needed to understand it fairly.

Too much context may overproduce unrelated material. Streaming communications often blend responsive, non-responsive, personal, privileged, confidential, and unrelated content. Broad production can raise proportionality, privacy, privilege, and confidentiality concerns.

That is why discovery teams increasingly need to think carefully about the proper unit of production.

Is it the individual message?

The thread?

A twelve-hour or twenty-four-hour window?

The document and its comments?

The workflow event?

The transaction sequence?

The answer will vary by matter, platform, and issue. But the question should be asked deliberately.

In modern discovery, the smallest available artifact is not always the fairest unit of meaning.

Implications for litigation partners

Rather than treating system-generated data as a purely technical issue, litigation partners and trial teams should embed the following practices directly into their case strategies.

• Expand preservation beyond emails: Include the full digital footprint—such as collaboration tools (Slack, Teams), mobile messaging, shared cloud documents, calendars, and system audit logs—before automated retention policies overwrite them.
• Interview for workflows, not just documents: Shift custodian interviews away from “Where are your documents?” to “How did this work actually happen?” to uncover how custodians interacted with systems, links, and automated notices.
• Define technical electronically stored information (ESI) protocols early: Proactively negotiate with opposing counsel to establish guidelines for context windows, message threads, edits, reactions, and hyperlinked cloud documents before disputes arise.
• Upgrade to connected review workflows: Abandon legacy “one-at-a-time” toolsets for platforms that can stitch together the full story, mapping chat messages to documents and system activity to specific users and timelines.
• Focus on context over collection: Do not waste resources trying to collect everything blindly. Target the specific surrounding data needed to build a verified timeline, test witness credibility, and defeat cherry-picked, out-of-context evidence.

From documents to decodable evidence

The practice of discovery is moving through a clear progression. First, litigation focused on the single document. Then, as communication fragmented, litigation had to account for multiple smaller documents and message parts.

Now, as those parts become smaller and more distributed, litigation increasingly requires attention to the underlying layer of system-generated data. That does not make traditional documents obsolete. Emails, PDFs, spreadsheets, contracts, and presentations still matter. But sometimes they must often be understood as part of a larger evidentiary environment. The modern challenge is not simply whether information exists. It is whether the information can be accessed, interpreted, tested, and used effectively.

As communication fragments, context becomes evidence. And in many cases, system-generated data is what makes the message understandable.

Closing thoughts: join the conversation

This is just one piece of the bigger conversation on the future of evidence. As legal professionals, we need to stay on top of emerging technologies.

Let’s continue the discussion on this LinkedIn post.