教程 11: Litigation Support & E-Discovery

Review Categories:
1. Responsive Documents (turn over)
2. Non-Responsive (don't turn over)
3. Privileged (attorney-client privilege)
4. Relevance: Key (directly supports case)
5. Relevance: Peripheral (tangentially related)
6. Relevance: Not Relevant (irrelevant)

I'm working on a product liability case (defective widget causing injury).
I need to categorize these discovery documents.

My review categories are:
- Responsive/Non-responsive
- Privilege status
- Relevance level (Key/Peripheral/Not Relevant)
- Contain admissions? (Yes/No)
- Timeline importance? (Yes/No)

Please analyze these documents and create a review matrix.

Analyzing document set for product liability case.

Document coding guidelines:
- Responsive: Contains information about widget design,
  testing, warnings, or injury circumstances
- Privileged: Internal communications with legal counsel
- Key Relevance: Direct evidence of defect knowledge
- Admissions: Statements acknowledging fault or danger
- Timeline Critical: Dates related to design/testing phases

[Analyze each document against these criteria]

I'm defending against an employment discrimination lawsuit (alleged race discrimination in termination).
Create a discovery review protocol that includes:

1. Five key document categories for this case
2. Definitions for "responsive" vs "non-responsive"
3. Privilege flags to watch for
4. Key relevance criteria
5. A decision tree for borderline documents
6. Quality control checkpoints

Make this practical for a team of 4 paralegals working 40 hours/week.

I have 250,000 documents from our ESI collection.
Many are duplicates or near-duplicates.

Please create a deduplication protocol that:
1. Identifies exact duplicates (same hash)
2. Finds near-duplicates (identical content, different formatting)
3. Handles similar emails in threads
4. Creates a priority ranking of which documents to keep
5. Estimates space savings

What percentage of legal document sets are typically duplicates?

Our ESI comes from these custodians:
- CEO (john.smith@company.com)
- CFO (jane.doe@company.com)
- Operations Manager (bob.jones@company.com)
- Legal Counsel (counsel@company.com)
- IT Director (tech@company.com)

Create a custodian profile showing:
1. Total documents from each person
2. Date ranges of their communications
3. Key recipients/senders per custodian
4. Potential privilege holders
5. Custodian-specific review focus areas

Bulk tag these documents:

For all emails from 2022-2024:
- Tag by custodian
- Tag by relevance (Key/Peripheral/Not Relevant)
- Flag if contains financial discussions
- Flag if contains product safety discussions
- Identify if responsive to these RFPs:
  * All documents re: product testing
  * All communications re: safety incidents
  * Budget documents for R&D department

Create a tagging matrix showing volume of documents
in each tag category.

Create a litigation hold notice and ESI preservation plan for:
- 8 custodians
- Email systems (Microsoft Exchange)
- Document management system (SharePoint)
- Database servers (SQL Server with daily backups)
- Mobile devices
- Cloud storage (OneDrive, Dropbox)

Include:
1. Detailed hold notice language
2. IT implementation steps
3. Cost estimates
4. Compliance verification checklist
5. Timeline with milestones

I'm analyzing the defendant CEO's deposition transcript (245 pages).

Please identify:
1. Key topics discussed (list with page numbers)
2. Topics where testimony was evasive
3. Topics with multiple contradictions
4. Topics with admissions or damaging statements
5. Topics that support plaintiff's narrative
6. Topics that support defendant's narrative
7. Unclear or confusing answers (with line numbers)

Format as a matrix so I can quickly locate relevant testimony.

The witness testified on page 45: "I never saw any safety concerns
about the product during my tenure."

But on page 187: "After we learned about the incident in March 2020,
we discussed internal safety issues."

Show me:
1. All other statements about safety concerns (with pages)
2. Timeline of when witness claims to have known about problems
3. Other potential contradictions in the testimony
4. Admission vs. denial statements

Create a "inconsistency report" that trial counsel can use
for cross-examination or impeachment.

Based on this deposition transcript, construct:

1. The witness's narrative: What story does their testimony tell?
2. Timeline of events per the witness
3. Admissions the witness made
4. Areas where testimony was weak or evasive
5. Testimony supporting plaintiff vs. defendant
6. Testimony about other witnesses
7. Testimony about documents they reviewed

Create a visual timeline showing dates mentioned in testimony.

[Upload a sample deposition transcript]

Analyze this deposition and provide:

1. Complete topic index with page references
2. Inconsistency report (contradictions within this deposition)
3. Witness credibility assessment (areas of evasion, confidence level)
4. Key admissions section
5. Opportunities for impeachment
6. Questions for follow-up depositions
7. Designations for trial (key testimony to use at trial)
8. Clips to exclude (damaging, unclear, hearsay)

Format this so a trial team can use it immediately.

From the deposition I just uploaded, identify:

1. Areas where witness contradicted themselves
2. Areas where witness contradicted documents
3. Areas where witness was evasive or used qualifiers
4. Areas where witness's knowledge seems incomplete
5. Areas that support plaintiff's theory of case

For EACH area, note:
- The testimony
- The contradiction/inconsistency
- The page numbers
- The witness's exact words (for quotation)

Based on [witness name]'s deposition, draft aggressive
cross-examination questions that:

1. Establish the contradiction
2. Narrow the witness's ability to explain
3. Use the witness's own words
4. Build on prior admissions
5. Set up for documentary evidence

Format each question:
Q: [Question]
Purpose: [What this establishes]
Support: [Page/testimony that backs this up]
Expected Answer: [What witness likely will/should say]
Follow-up if evasive: [Aggressive follow-up]

Organize these 50 cross-examination questions by:

1. Timeline/Chronology questions
2. Knowledge/Notice questions
3. Documents and emails
4. Prior statements and inconsistencies
5. Industry standards questions
6. Motive/Bias questions
7. Admission questions
8. Impeachment questions

For trial, what order should these be asked?
What's the strategic flow?

Given this deposition testimony, create a complete
cross-examination outline including:

1. Opening statement setting up major inconsistency
2. Questions establishing foundation
3. Questions using witness's own words
4. Questions about documentary evidence
5. Questions about prior statements
6. Questions establishing contradictions
7. Closing summary question

Include witness designations showing where testimony
is located in the video deposition.

From all documents and depositions, extract:

1. Event date
2. Document date (when created)
3. Type of event (Internal discussion, External communication,
   Physical event, Decision point)
4. Key parties involved
5. Document IDs/citations
6. Brief description
7. Significance (Background, Key Event, Turning Point, Damaging, Exculpatory)

Format as spreadsheet with sortable columns.

Create a comprehensive case timeline showing:

1. All events in chronological order
2. Key decision points
3. When key players knew what information
4. Document production deadlines in the discovery sequence
5. Litigation milestones
6. Business events impacting the case
7. When warnings/notices should have been given

Include:
- One-page executive timeline
- Detailed timeline (page per month)
- Visual timeline graphic
- Decision-maker knowledge timeline

Based on the events timeline, create:

1. Plaintiff's preferred timeline narrative
   (what sequence of events supports plaintiff)
2. Defendant's preferred timeline narrative
   (what sequence supports defendant)
3. Areas where parties disagree on dates
4. Timeline holes (missing information)
5. Timeline inconsistencies (who says what happened when)

For each narrative, identify the 5 most important dates.

[Upload multiple documents from various dates, plus
deposition transcripts with testimony about dates]

Build a complete case timeline that:

1. Extracts all dates from all sources
2. Identifies conflicts between document dates and testimony
3. Creates a consolidated timeline
4. Marks gaps in documentation
5. Shows decision-maker awareness over time
6. Highlights "should have known" dates
7. Includes litigation milestones

Show timeline in three formats:
- Spreadsheet (sortable)
- Narrative (month-by-month)
- Visual (graphic representation)

I have statements from 8 witnesses in this case.

For EACH witness statement, identify:

1. Key facts claimed by witness
2. Contradictions with other witness statements
3. Contradictions with documents
4. Admissions in the statement
5. Areas of personal knowledge vs. hearsay
6. Witness credibility factors
7. Corroborating documents
8. Areas where witness is vulnerable

Create a witness comparison matrix.

Compare witness statements to:

1. Documents produced
2. Email communications
3. Photographs/physical evidence
4. Other witness statements
5. Deposition testimony
6. Prior statements or interviews

Flag:
- Where evidence corroborates witness
- Where evidence contradicts witness
- Where evidence is missing
- Where witness fills gaps in documentation

For each witness, assess:

1. Internal consistency (contradictions within their own statement)
2. External consistency (contradicts other evidence)
3. Specificity of testimony (vague vs. detailed)
4. Motive/bias factors
5. Prior criminal history or credibility issues
6. Demeanor indicators (from deposition video if available)
7. Areas of evasion
8. Strong vs. weak parts of testimony

Assign credibility score: Strong / Moderate / Weak

[Upload 4-5 witness statements]

Analyze and produce:

1. Witness matrix comparing key facts
2. Inconsistency report with page references
3. Document corroboration table
4. Credibility assessment for each witness
5. Critical gaps where evidence is missing
6. Recommended deposition questions
7. Trial designation strategy (which witnesses to call)
8. Impeachment opportunities

Create search protocols for evidence items:

For product liability case, I need to find evidence about:
1. Product design and testing
2. Prior incidents/complaints
3. Cost-benefit analyses
4. Safety warnings and labels
5. Competitor products
6. Industry standards
7. Executive knowledge of risks
8. Post-incident remediation

For EACH category:
- Key search terms
- Document types to focus on
- Date ranges
- Custodians most likely to have
- Boolean search strings
- Expected document counts

During depositions, I need to rapidly locate:
1. Specific documents witness is testifying about
2. Documents that contradict testimony
3. Documents witness claims not to know about
4. Email chains referenced in testimony
5. Specific dates/time periods discussed

Create a "rapid retrieval system" that lets me:
- Search by document ID
- Search by date
- Search by keyword
- Search by custodian
- Get results in seconds
- Pull to screen instantly

What technology solutions would enable this?

During trial, I need immediate access to:
1. Exhibits referenced in testimony
2. Prior inconsistent statements
3. Testimony from other witnesses about same topic
4. Deposition designations about this evidence
5. Expert reports addressing this evidence

Create a trial evidence management system including:
- Quick search capability
- Hot-key access to frequently used exhibits
- Visual display system
- Citation/reference system
- Organization by case theory/theme

Create a complete evidence search and retrieval protocol for:
1. Pre-trial (office research)
2. During deposition (in-person)
3. During trial (courtroom)

Include:
- Search term library (organized by topic)
- Boolean search strings
- Database access procedures
- Mobile/laptop requirements
- Quality control (verify correct document)
- Citation/reference system
- Backup systems if technology fails

Identify communications that may be privileged:

1. Attorney-Client Privilege
   - Who is the attorney/client?
   - Is there a business purpose exception?
   - Are non-lawyers present?
   - Was privilege waived?

2. Work Product Doctrine
   - Created in anticipation of litigation?
   - Created by attorney or at direction of attorney?
   - Opinion vs. fact work product?

3. Common Interest Privilege
   - Multiple parties' common legal interest?
   - Proper agreement in place?

4. Expert Privilege
   - Consulting expert (work product)?
   - Testifying expert (some protection)?

Create a privilege decision tree.

Flag communications with:
1. Distribution to legal counsel
2. Subject lines mentioning "legal," "attorney," "counsel"
3. Email addresses of known attorneys
4. Confidentiality markings
5. References to "attorney," "legal advice," "legal review"
6. References to pending/threatened litigation
7. Explicit privilege assertions

Create a database field: "Privilege Flag: Yes/No/Maybe"
For "Maybe" items, require attorney review.

For withheld documents, create privilege log showing:

1. Document ID
2. Document date
3. Author
4. Recipients
5. General description
6. Type of privilege asserted
7. Basis for privilege claim
8. Privilege duration

Format:
- Sortable spreadsheet
- Organized by document date
- Cross-indexed to responsive documents

Create a privilege review protocol for your discovery set that includes:

1. Privilege identification flowchart
2. Privilege decision rules (when to withhold)
3. Metadata fields to check
4. Privilege log template
5. Quality control procedures
6. Common mistakes to avoid
7. Timeline/staffing requirements
8. Cost estimates

Assume 50,000 documents and use your historical privilege incidence rate.

I have 20 RFPs (Requests for Production). For each one,
define "responsive" documents:

RFP 1: "All documents regarding product testing"
Responsive = Any document mentioning testing, QA, quality control
Non-responsive = Marketing materials (unless discussing testing)

RFP 2: "All communications with [Company X]"
Responsive = Any email or document with [Company X]
Non-responsive = Emails about [Company X] but not to/from them

For EACH RFP:
- Exact language
- Definition of responsive
- Definition of non-responsive
- Ambiguous categories (coding guidance)
- Examples of responsive docs
- Examples of non-responsive docs

Create a document review coding form with:

1. Document ID
2. RFP 1 responsive? (Yes/No/Privileged/Objection)
3. RFP 2 responsive? (Yes/No/Privileged/Objection)
4. ... [continue for all RFPs]
5. Privilege assertion (if applicable)
6. Reviewer name and date
7. Confidence level (certain / confident / uncertain)
8. Notes/issues

Make this form usable for paralegals with
legal assistant training (not law students).

Build a QC process that:

1. Re-reviews 5% of documents coded "responsive"
2. Re-reviews 10% of documents coded "privileged"
3. Tracks agreement rates by reviewer
4. Identifies reviewers with coding errors
5. Provides corrective feedback
6. Escalates difficult decisions to attorney

Set target agreement thresholds based on your court expectations and case risk profile.
Set QC staffing/time budgets based on pilot variance and reviewer calibration results.

[Provide sample RFPs]

Build a complete responsive document coding system:

1. Decode each RFP into clear coding rules
2. Create decision tree for ambiguous documents
3. Draft coding form/worksheet
4. Develop QC protocol
5. Create training materials for coders
6. Estimate time/cost for 25,000 documents
7. Build escalation procedure for difficult documents
8. Create privilege subset process

Show how this integrates with your overall discovery workflow.

I need to understand email communications in this case.

For all emails, identify:
1. Complete threads (original + all replies)
2. Thread participants (all senders/recipients)
3. Thread timeline (dates of messages)
4. Forward chains (emails forwarded to new recipients)
5. Re-introduced threads (same subject, new conversation)
6. Custodians involved
7. Critical information in threads
8. Contradictions within threads

Create a thread catalog.

For the 50 most important email threads:

1. Identify the core issue being discussed
2. Map how the discussion evolved
3. Show who said what and when
4. Identify decisions made
5. Identify admissions
6. Identify contradictions
7. Show any changes in narrative
8. Identify key missing replies (someone should have responded but didn't)

Create visual maps of the threads.

Across all email communications, analyze:

1. Communication frequency between key parties
2. Timing of communications (urgent vs. routine)
3. Escalation patterns (who gets copied when)
4. Information flow (who tells whom)
5. Knowledge distribution (who knew what when)
6. Decision-making patterns (how decisions made)
7. Problem-solving approach (how issues addressed)
8. Tone/demeanor changes over time

What communication patterns support/undermine your case?

[Upload email set with 100+ messages in interconnected threads]

Provide:

1. Complete thread inventory with dates and participants
2. 10 most important threads - detailed analysis of each
3. Communication pattern analysis (who talks to whom)
4. Timeline showing when communications occurred
5. Admissions/damaging statements (with citations)
6. Visual thread maps for key discussions
7. Knowledge timeline (who knew what when)
8. Missing communications (gaps in expected email)

Format for use in trial preparation and expert reports.

Find all documents where:
- Someone expresses concern about product safety
- The concern is about our product, not competitors
- The communication is between company executives
- The date is before launch

Find documents that show:
1. Someone at the company knew about a problem with [Product X]
2. They knew before [Date]
3. They didn't fix it (or delayed fixing it)
4. The problem later caused harm (per incident reports)
5. They didn't warn customers

Show me:
- Which executives knew
- When they knew
- What they did/didn't do about it
- What harm resulted
- Evidence about knowledge of harm

Find documents matching any of these patterns:

Pattern A: Executive + safety concern + before launch
Pattern B: Email exchange + disagreement + safety topic
Pattern C: Document + test results + concerning findings
Pattern D: Communication + cost-benefit + safety tradeoff
Pattern E: Incident report + similar prior incident

Show results organized by pattern,
with relevant excerpts highlighted.

I need to find evidence of four key claims:

1. Company knew about the defect
2. Company knew it was dangerous
3. Company chose profit over safety
4. Company knew of similar incidents

For EACH claim, create:
- Plain English query description
- Expected types of documents
- Key phrases/terms to search for
- Boolean search terms as backup
- Results (with document locations)

Use natural language phrasing, not technical boolean.

I have 15,000 discovery documents.

Cluster them by analyzing:
1. Subject matter (product design, testing, incidents, communications, etc.)
2. Time period
3. Participants/custodians
4. Document type (emails, reports, spreadsheets, etc.)
5. Similarity of content and language patterns

Show me:
- Number of natural clusters
- Size of each cluster
- Key topics in each cluster
- Which documents are bridges between clusters
- Outliers that don't fit main clusters

Based on the document clusters, organize by legal themes:

1. Theme 1: Knowledge of Defect
   - Documents showing awareness of problem
   - Timeline of knowledge
   - Who knew what

2. Theme 2: Failure to Warn
   - Documents about warning decisions
   - Cost-benefit analyses
   - Warning label discussions

3. Theme 3: Prior Incidents
   - Reports of similar problems
   - Complaint tracking
   - Follow-up actions

4. Theme 4: Concealment
   - Documents destroyed?
   - Information withheld?
   - Contradictory narratives

For each theme, show key documents in order.

My expert will testify about:
- Industry standards for product testing
- Adequacy of warnings
- Comparison to competitor products

Cluster documents that:
1. Show your product testing vs. industry standard
2. Show warning adequacy vs. competitors
3. Show decision-making about safety
4. Show cost-benefit tradeoffs
5. Show knowledge of risks

Organize clusters so expert can cite them directly.

[Upload 50-100 documents from a litigation matter]

Perform complete clustering analysis:

1. Identify natural clusters in documents
2. Analyze cluster themes
3. Show bridge documents
4. Identify outliers
5. Create thematic organization
6. Link to case theories
7. Recommend key documents for each expert
8. Create visual cluster map

Show how document clustering reveals hidden patterns
in your discovery set.

Create a review protocol for [case type] that defines:
- Responsive vs. non-responsive
- Privilege categories
- Relevance levels
- Admissions and timeline importance

Analyze this deposition transcript and:
- Identify all topics discussed
- Flag contradictions and inconsistencies
- Note admissions and damaging statements
- Draft cross-examination questions

Extract all events and dates from these documents and depositions.
Create timeline showing:
- Chronological events
- Key decision points
- Knowledge timelines
- Strategic turning points

Analyze these emails as threads showing:
- Complete thread conversations
- Participant communications
- Information flow over time
- Critical admissions or contradictions

Compare these witness statements identifying:
- Contradictions between witnesses
- Contradictions with documents
- Admissions and credibility issues
- Gaps in evidence

Flag potentially privileged communications:
- Attorney-client privilege
- Work product doctrine
- Common interest privilege
- Create privilege log with descriptions

Find evidence of [specific claim] showing:
- Who knew what
- When they knew it
- What they did about it
- How this harms [side]