Forbidden Phrases

Phrases that should not appear in an academic submission cover letter. The regex patterns in presubmission_check.py are the single source of truth; this file is the human-readable mirror. Each tier maps to a code group the script emits:

Tier	Script constant	Codes	Trigger
1 (openers)	LETTER_OPENER_CLICHES	L2a-L2e	first content line only
2 (marketing adjectives)	BANNED_TONE_PATTERNS	AI1-AI14	3+ occurrences
3 (cover-letter clichés)	LETTER_BANNED_PHRASES	J1a-J1i	1+ (major at 3+)
4 (generic-fit)	LETTER_GENERIC_FIT_PHRASES	J4a-J4d	1+, with replacement hint

Tier 1: Low-effort openers (always flag)

These signal a low-effort letter to top-journal editors:

• "We are pleased to submit"
• "We are excited to share"
• "We are delighted to submit"
• "We are honored to submit"
• "It is our great pleasure to submit"
• "Enclosed please find"
• "Please find enclosed"
• "Please find attached"
• "We hereby submit"

Tier 2: Marketing adjectives (flag at 3+ occurrences)

These inflate without adding evidence. Inherits from paper-audit BANNED_TONE_PATTERNS:

• innovative
• pioneering
• revolutionary
• transformative
• breakthrough
• unprecedented
• remarkable
• superior
• state-of-the-art
• "highlights the potential of"
• "paves the way"
• "profound challenges"
• "at its essence"

Tier 3: Cover-letter-specific banned phrases (flag at 1+ occurrence)

These are template-style fillers that editors at top journals discount:

• "novel and innovative"
• "groundbreaking"
• "first-of-its-kind"
• "game-changing"
• "paradigm shift"
• "cutting-edge"
• "of great interest"
• "will be of broad interest"
• "the field is in need of"

Tier 4: Generic-fit phrasings (flag with suggested replacement)

These indicate the author has not read the journal carefully:

• "your journal" / "your prestigious journal" — name the journal explicitly
• "fits well with the scope" — name the specific scope dimension
• "broad readership" — name the specific reader profile
• "important contribution to the field" — name the contribution category

What is acceptable

• Confident factual statements: "Our method reduces inference latency by 47% on the ImageNet validation set."
• Bounded comparisons: "This extends the framework of Chen et al. (2024) by introducing X."
• Specific journal connections: "This work directly addresses the open question raised in [recent paper from the target journal]."

Why these are forbidden

The pattern across these tiers: they substitute marketing for evidence. An editor reading 50 letters a week can spot template-language instantly and downgrades the letter's signal value. Cover letters that pass presubmission_check.py with zero Tier 1-3 hits read as written-with-attention.