What a dead zone looks like

Dead zones take many shapes. They include exposition-heavy passages that repeat known facts, scenes that linger on low-stakes activity, shifts into irrelevant subplots, prolonged internal monologue that adds little, and transitional chapters that exist only to move characters from A to B. Dead zones can also be structural: an act that runs long without a clear escalation, multiple scenes that cover the same emotional territory, or a middle section where stakes feel reduced. The common outcome is the same. Reader attention drops and narrative urgency fades.

Why finding dead zones matters

Of course, this depends on genre norms. A horror or thriller writer has different requirements to an historical novelist. Removing or reworking dead zones improves momentum, heightens reader investment, and increases the perceived pace without rushing character development. Tight pacing clarifies which scenes are essential, which can be merged, and which should be refocused. Faster, sharper drafts are easier to sell to agents and editors because they demonstrate structural control and respect for the reader’s time.

How AI diagnoses dead zones

AI can analyze your manuscript at scale and surface patterns that indicate slow areas. It can:

• Measure sentence length and variation to reveal rhythmic flattening.
• Identify chapters or scenes with low action verbs and high rates of exposition words.
• Flag repeated themes, facts, or imagery that suggest redundancy.
• Produce a chapter-by-chapter urgency score based on conflict density, stakes language, and beat frequency.
• Highlight long stretches without a visible external goal or scene-level turning point.
• Compare pacing profiles across POVs to find imbalance and uneven chapter pacing.

AI’s strength is pattern recognition across thousands of words, which makes it efficient at pointing you to likely dead zones for human judgement.

Signs AI will flag as high risk

• Long average sentence length combined with low dialogue ratio in multiple consecutive chapters.
• High density of “telling” verbs and abstract nouns versus concrete actions.
• Repeated factual statements or backstory fragments across chapters.
• Extended scenes where the protagonist’s external goal is unclear or absent.
• Low incidence of decision points or escalating consequences.
• Abrupt tonal shifts that break momentum without a purposeful scene purpose.

Workflow to diagnose and fix dead zones

1. Prepare and back up the manuscript before any automated pass.
2. Run a pacing diagnostic on chapter-sized chunks rather than the whole manuscript at once.
3. Review flagged sections and annotate why the AI marked them: redundancy, low stakes, exposition, or lost external goal.
4. Triage flagged scenes into three buckets: cut, condense, or refocus.
5. For the cut pile: remove scenes that duplicate information or serve only as filler; merge necessary beats into surrounding scenes.
6. For the condense pile: shorten description, collapse repeated beats, and trim internal monologue to the elements that reveal character or advance plot.
7. For refocus pile: identify and amplify a clear external goal, increase conflict, add concrete sensory details tied to stakes, or introduce a ticking constraint.
8. Re-run the pacing diagnostic on revised sections to confirm improvement.
9. Perform a read-aloud pass to assess rhythm and energy after revisions.
10. Repeat until chapter-by-chapter urgency scores show more even rises and dips appropriate to story structure.

Line-level fixes to speed pace

• Replace passive or abstract verbs with concrete, active verbs that show motion or choice.
• Convert long paragraphs of summary into short scenes or micro-scenes with sensory detail and a small decision.
• Break long sentences into varied beats to reintroduce momentum and make beats easier to scan.
• Cut repeated physical or emotional descriptions that do not add new information.
• Swap expository blocks for dialogue or action that reveals the same facts through conflict or character choice.
• Introduce a small constraint or deadline in the scene to create immediate pressure.

Scene-level strategies to sharpen focus

• Give each scene a single, clear visible goal for the POV character and a measurable success or failure condition.
• Raise the cost of failure for the character to increase urgency.
• Keep scenes short unless the extended length serves a deliberate emotional or structural purpose.
• Use micro-conflicts inside scenes: interruptions, misunderstandings, timing mishaps, or resource scarcity.
• End scenes with a new complication or question that forces the reader to turn the page.

Prompts and templates you can reuse

• Diagnostic prompt: “Analyze this chapter for pacing. Return a score from 1 to 10 for urgency and list up to five reasons the score is low, prioritized by impact.”
• Redundancy detector: “Flag sentences or paragraphs that repeat facts introduced earlier in the manuscript and suggest concise alternatives or merge points.”
• Action density check: “List every sentence with an explicit external action verb and report gaps longer than X words without an action.”
• Scene refocus template: “Here is a scene. Identify the visible external goal, list three ways to raise the stakes, and propose two concise versions of the scene: one condensed, one refocused with a stronger goal.”
• Read-aloud rhythm prompt: “Convert this paragraph into three beats suitable for vocal performance and recommend where to break sentences to improve rhythm.”

Examples of practical changes that work

• Condense a ten-paragraph room-description into one paragraph that reveals a single sensory detail linked to a character’s emotion and move the rest to a file for later use.
• Replace an internal monologue that explains a character’s decision with a short scene showing the decision being made under pressure.
• Merge two consecutive low-conflict scenes by preserving the most consequential beats and dropping redundant exposition.

When not to cut for pace

• Scenes that exist to slow pace intentionally for emotional digestion or thematic contrast.
• Extended passages that develop essential character empathy or that prepare the reader for a tonal shift.
• Experimental or lyrical sequences where rhythm and repetition are purposeful artistic choices.

Final checklist before submission

• Do all scenes have a visible external goal or a clear reason for their length?
• Are repeated facts and backstory consolidated and presented only where they serve tension or character change?
• Does the middle maintain escalating stakes or deliberate reversals to avoid sagging?
• Are sensory details aligned with the POV character and used to heighten immediacy?
• Have you re-run diagnostics and read the revised sections aloud to confirm regained momentum?

Closing guidance

Use AI to map the problem and give options, then apply human judgement to decide which fixes preserve emotional truth while restoring momentum. Treat the AI as a diagnostic tool and an honest first-draft editor, not the final arbiter. When you combine automated pattern detection with targeted rewrites and read-aloud validation, you turn slow stretches into purposeful beats and keep readers moving through your story.

About the author

As an introvert haunting the corners of storytelling festivals, it’s incredibly difficult to track Emma down. She’s best known for writing Scottish fiction about working-class women and communities and their misrepresented lives. You can find her recent book The Secret Cult of the Miners’ Library here. Or get writing help here.

L’article Use AI to Find Dead Zones in Your Manuscript and Speed Up Pacing est apparu en premier sur Emma Parfitt Proofreading Editing Services.

Why run an AI line-edit before submission

According to AI it excels at consistent mechanical fixes, catching patterns human eyes miss, and producing fast alternatives for awkward phrasing. That makes it ideal for a final sweep to remove typos, tighten language, and flag inconsistencies. Um … not in my experience. As AI is only a tool: it cannot fully replace a careful human pass or editorial judgement. The checklist below will help you keep speed without sacrificing voice, accuracy, or originality.

The checklist

(run these in order)

• 1. Global read-through for tone drift Ask the AI to compare the start, middle, and end of your manuscript for consistency in tone and register, and to flag chapters where voice or tense shifts unintentionally.
• 2. Dialogue distinctiveness scan Have the AI list patterns for each character: average sentence length, common words or phrases, and register. Flag characters whose speech metrics overlap too closely.
• 3. Mechanical clean-up (grammar, punctuation, spelling) Run a constrained pass that fixes only objective mechanical errors. Tell the AI explicitly not to alter slang, dialect, or deliberate grammatical choices.
• 4. Passive voice and weak verb report Request a list of instances of passive voice and soft verbs (e.g., “was,” “had,” “felt”) with suggested stronger verbs or active constructions, focusing on scenes that should carry momentum.
• 5. Repetition and filler detection Ask the AI to point out repeated words, phrases, and filler constructions within each chapter and to suggest concise alternatives or indicate where repetition is purposeful.
• 6. Head-hopping and focalization check Have the AI identify sentences or paragraphs that shift perspective within a scene and mark them for you to confirm whether the shift is deliberate.

And yes we keep going!

• 7. Continuity and timeline audit Ask for a concise timeline of major events and character states across chapters; flag inconsistencies in dates, ages, or physical states.
• 8. Named-entity and prop consistency Request a list of character names, locations, objects, and their first mention; flag inconsistent spellings, changing traits, or props that vanish and reappear.
• 9. Sensory balance and specificity pass Ask the AI to score scenes for sensory detail (sight, sound, smell, touch, taste) and suggest one concrete sensory detail per under-specified scene aligned to the POV character’s life experience.
• 10. Red-flag content scan Request detection of potentially problematic content: stereotypes, uncontextualized trauma, or factual red flags that may need sensitivity reading or expert review.
• 11. Query/package readiness check Have the AI produce a short synopsis, 1–2 sentence hook, and a 250–300 word pitch for submission, then compare tone to the manuscript to ensure alignment.
• 12. Final voice-preservation pass Ask the AI to propose only minimal edits that preserve sentence rhythm and diction; produce two variants per flagged sentence: “clean” and “voice-first.”

How to prompt AI for safe, useful edits

• Start with explicit constraints: “Edit for clarity and mechanical correctness only; do not change dialect, slang, sentence fragments, or repeated motifs used for voice.”
• Provide voice cues: “Voice: sardonic first person; short, clipped sentences; recurring nautical metaphors.”
• Ask for labeled outputs: “Return edits as: 1) Original sentence, 2) AI-clean suggestion, 3) Voice-first suggestion, 4) Short rationale (10–15 words).”
• Use small batches: submit chapter-sized excerpts or 500–1,000 word chunks to maintain context and control.
• Request a changelog: ask the AI to produce a numbered list of every edit type it applied (grammar, tense, wording) so you can quickly review categories rather than individual changes.

Practical workflow to combine AI with human judgment

1. Backup your manuscript and save a timestamped copy before any AI pass.
2. Run the constrained mechanical clean-up first and accept changes only where they don’t alter voice.
3. Run the repetition, passive-voice, and continuity checks and mark suggested fixes in a separate file or using tracked changes.
4. Do a focused voice-first pass: accept only the “voice-first” variants that preserve cadence; reject or adapt others.
5. Perform a read-aloud pass—either by voice actor, text-to-speech, or out-loud yourself—to catch rhythm, tone, and dialogue authenticity.
6. Send key chapters to a trusted human editor or beta reader with notes listing AI changes so reviewers focus on voice, character, and emotional impact rather than copy edits.
7. Implement final human edits, run one last mechanical check, then prepare submission package materials.

Red flags and when to skip AI for a section

• If scenes hinge on cultural nuance, lived trauma, or technical accuracy, prefer human experts or sensitivity readers first.
• If you rely heavily on unique diction, dialect, or experimental syntax, avoid broad AI line-edits that lack voice scaffolding.
• If your draft contains contractual obligations or third-party materials you cannot risk exposing, don’t upload those sections to public/free AI tools.

Quick templates you can reuse

• Mechanical clean-up template: “Fix grammar, punctuation, and spelling only. Do not change slang, intentional sentence fragments, or character-specific diction. List edits with short rationale.”
• Voice-first template: “Edit for clarity while keeping [character name]’s voice. Voice cues: [3–4 keywords]. Provide two variants per sentence: clean and voice-first.”
• Continuity audit template: “Create a timeline of key events and list all named characters, locations, and props with first-mention chapters. Flag inconsistencies.”

Final thoughts

An AI line-edit should leave your manuscript cleaner but unmistakably yours. Use the checklist to structure passes, protect character voice, and combine machine speed with human discernment. When AI is confined by smart prompts, strong templates, and a human-in-the-loop workflow, it becomes a powerful editing assistant that helps you submit work that’s polished, faithful to your voice, and professionally ready.

About Emma

As an introvert haunting the corners of storytelling festivals, it’s incredibly difficult to track Emma down. She’s best known for writing Scottish fiction about working-class women and communities and their misrepresented lives. You can find her recent book The Secret Cult of the Miners’ Library here. Or get writing help here.

L’article The AI Line-Edit Checklist est apparu en premier sur Emma Parfitt Proofreading Editing Services.

Why AI edits can hurt character voice

AI models are optimized for general clarity and broad readability. They favor neutral, idiomatically safe constructions that minimize ambiguity. That optimization produces consistent, polished prose but also flattens stylistic extremes. Unique diction, syntactic quirks, dialect, register shifts, purposeful repetition, and deliberate “errors” used for characterization. Important author tools, in other words. Can all be flagged as problems and replaced with safer alternatives by an AI assistant.

AI also lacks lived experience and the subtle sense of motive and embodied perspective that shapes authentic voice. It cannot reliably infer why a narrator shortens sentences, misuses an idiom, or shifts tense for psychological effect. When an AI “fixes” these features it substitutes a model of clarity for a model of character, and you lose the cues readers need to inhabit the mind behind the words.

Common patterns of voice erosion and how to spot them

• Over-correction to neutrality AI replaces idiosyncratic phrasing with bland alternatives. Spot it by comparing problematic passages to earlier drafts or to dialogue samples where the voice feels strong.
• Regularized syntax AI rewrites sentence fragments, run-ons, or elliptical constructions into standard grammar. Spot it by listening for lost rhythm or a change in sentence length distribution.
• Tone leveling AI smooths out registers so distinct characters sound the same. Spot it by reading only dialogue lines in sequence and checking whether each character still “speaks” differently.
• Loss of purposeful “mistakes” AI corrects nonstandard grammar, slang, or malapropisms that were intentional. Spot it by marking places where language choices inform class, education, or mindset.
• Diminished sensory specificity AI substitutes generic adjectives and erases odd but revealing metaphors. Spot it when imagery becomes recyclable or when images stop aligning with the character’s lived experience.

Fast line-level fixes you can use immediately

1. Re-run the original line through your voice template. Remind the AI of tone cues like wry, forgiving, childlike, or brusque, and ask for alternatives rather than single rewrites.
2. Use micro-examples. Show the AI one or two lines that exemplify the voice you want and ask it to match them.
3. Mark “do not change” elements inline. Bold or tag proper nouns, slang, and sentence fragments you want preserved before asking for edits.
4. Ask for variants: request one “AI-clean” edit and one “voice-preserving” edit so you can choose.
5. Keep an eye on rhythm. If sentence length changes noticeably after an edit, restore the original cadence by reintroducing shorter or longer beats.

Scene and character-level strategies

• Create a one-page voice profile for each point-of-view character. Include keywords, rhythm notes, favored imagery, common sentence patterns, and typical grammatical quirks. Paste the profile into prompts whenever you ask an AI to edit a scene.
• Build a short list of “signature phrases” for each character and keep a find-and-replace checklist to prevent the AI from overwriting them.
• Use comparative prompts that ask the AI to edit for clarity but preserve the protagonist’s sensory field and worldview. For example, ask it to “keep imagery grounded in the protagonist’s trade and hometown” or to “preserve emotional undercutting by using understatement.”
• Reserve AI for neutral tasks like consistency checks, easy fact lookups, line edits for grammar, and pacing diagnostics. Keep creative choices — metaphor, rhythm, unusual syntax — to humans. Or to an AI only when guided by strong voice scaffolding.

Recovery workflow when AI has already eroded voice

1. Retrieve the pre-AI draft or the last manual revision and put the two versions side by side.
2. Identify the key voice losses: diction, syntax, imagery, or tone. Mark examples in both versions.
3. Restore the original phrasing selectively where voice matters most, such as chapter openers, internal monologue, and pivotal dialogue.
4. Use a “voice-first” pass: ask the AI only to flag mechanical errors while leaving flagged stylistic elements untouched.
5. Run a focused read-aloud pass to catch rhythmic changes that slip past silent reading. Read characters’ dialogue in different voices to ensure distinctness.
6. Recruit at least one human reader who knows the manuscript and can point out where characters sound the same or where the narrator feels flattened.

Prompts and template examples for voice-safe edits

• Create a short template to preserve voice: “Edit for clarity and grammar but maintain this character’s voice. Voice cues: [two to three keywords]. Keep sentence fragments, slang, and [signature phrase]. Output two variants: 1) minimal tidy; 2) voice-forward rewrite with one optional smoother sentence.”
• Diagnostic prompt: “Highlight up to ten edits that change diction, rhythm, or register. For each, explain why it harms character voice and suggest an alternative that preserves voice.”
• Restoration prompt: “Here is a paragraph from draft A and the AI-edited paragraph from draft B. Produce a merged paragraph that keeps the most distinctive words and rhythm from draft A while applying only essential mechanical corrections.”

Practical checklist before accepting AI edits

• Does the edit preserve the narrator’s sensory field and primary metaphors?
• Would this sentence still make sense if read by the character out loud?
• Is any slang, dialect, or grammatical irregularity explained by social background, education, or emotional state?
• Did the AI remove repetition or eccentric phrasing that was serving a stylistic or narrative purpose?
• After the edit, do all characters still have unique speech patterns?

Parting words

AI is a powerful drafting and editing ally when used sparingly. Protect your voice with profiles, signature phrases, and targeted prompts. Treat AI edits as suggestions, not a final authority on your book. Reclaim rhythm with read-alouds and a human eye. When you make these checks habitual you gain speed without paying the price of uniformity, and you preserve the singular human presence at the heart of your novel.

About Emma

As an introvert haunting the corners of storytelling festivals, it’s incredibly difficult to track Emma down. She’s best known for writing Scottish fiction about working-class women and communities and their misrepresented lives. You can find her recent book The Secret Cult of the Miners’ Library here. Or get writing help here.

L’article When AI Suggests Edits That Hurt Character Voice and How to Fix Them est apparu en premier sur Emma Parfitt Proofreading Editing Services.

Respect copyright and licensing

Choose AI tools that state clearly what training data they use and what rights they claim over user content. Prefer services that explicitly say you retain ownership of what you upload. Plus that they will not use your text to train models without consent. When a tool offers paid or enterprise tiers labeled non‑training or private, read the contractual language. This will confirm their limits on reuse, sublicensing, and commercialisation. Keep a dated local backup of drafts and snapshots of the tool’s Terms of Service. Important: at the time you uploaded material, so you can prove provenance if needed.

Be transparent with collaborators

Before work begins, document how AI will be used across drafting, editing, and marketing stages and get written agreement from all coauthors and contributors. Specify whether AI assists only with mechanical edits or whether it will generate creative passages, and agree how that affects authorship credit, copyright ownership, and royalties. Set boundaries for what each collaborator may upload to third‑party tools and create a shared folder or log that records AI prompts, outputs used, and who approved them. Revisit and update the agreement at major milestones such as first draft completion, agent submission, or contract offers.

Protect privacy

Treat unpublished manuscripts, private correspondence, and third‑party source material as sensitive data and avoid pasting them into free or unclear AI services. Use tools that guarantee data isolation or choose local/offline models for anything you cannot risk being stored or reused. When you must share private material with editors, beta readers, or sensitivity readers, get explicit consent that they will not feed the text into external AI services and ask them to confirm deletion after review. Anonymize or redact identifying details when you need remote help on vulnerable or legally sensitive scenes.

Credit human contributors

Acknowledge editors, line editors, developmental editors, sensitivity readers, fact‑checkers. Basically, any collaborator whose input materially shaped the manuscript. State in your acknowledgements or contributor notes what role each person played, and be transparent about where AI tools were used and for what purpose. If contributors worked under contract, ensure payment and credit terms reflect the scope of their work and whether they used AI tools themselves. Treat human labor as irreplaceable expertise and compensate or credit it accordingly rather than substituting AI for professional judgment.

Basic takeaways

Use AI to augment creative work, not to replace the author’s judgment, voice, or ethical responsibility. Check out my other articles about legal rights and AI use.

About the Author

As an introvert haunting the corners of storytelling festivals, it’s incredibly difficult to track Emma down. She’s best known for writing Scottish fiction about working-class women and communities and their misrepresented lives. You can find her recent book The Secret Cult of the Miners’ Library here. Or get writing help here.

L’article The Ethical and legal considerations of writing with AI assistance est apparu en premier sur Emma Parfitt Proofreading Editing Services.

Safe uses

Or safe-ish? A brainstorming tool for character prompts, setting ideas, plot twists, and scene starters if you are stuck. One example is plotting. When you have a novel idea you can draft a complete outline to save time, then write it out scene by scene adding your own ideas and style. Then, if anything like me, completely changing it as you go along! Keep your original AI output as proof that you have completely changed and made it into your own.

Use it for those difficult sentences that you think could be better phrased. For example to pull up 10 alternative sentences, even more! Then sit back and read through them and tweak as needed.

The same is true for structural analysis. This involves sorting out pacing problems, revealing weak or repetitive phrasing. Though ensure you tell it the exact genre and audience, or the output is generic. Plus, if you use dialects and local words it tends just to be annoying, unless you tell it to ignore them.

This tip might be in the unsafe category as AIs make mistakes. But you can use it to check historical facts as a shortcut to trawling through internet sites. Using AI as a search assistant you can summarize background material, compile quick facts, and produce concise timelines. Though I think historical bloggers and folks with passion, sharing their experience online would appreciate direct links, comments and reads!

Sort out those mechanical, time sucking edits. This includes grammar, punctuation, consistency checks, and basic line edits to save time before human revision. However, it does make mistakes and has a nasty habit of rewriting what you don’t want and introducing new errors. So, for a professional finish hire a proofreader.

Have you heard of prompt templates? These are instructions capture what you want the AI to do, the tone you want, constraints, and examples that instruct the AI to follow each time. It means any resulting text better matches your voice and constraints. For instance, do you want UK or US English? Are you writing horror or young adult fiction?

Type up your blurb/synopsis ideas for the back of the book and ask the AI to create it for you using your text. Use more detailed instructions for more specific results. I tested AI for this, but found it oversimplified the text and introduced phrasing I didn’t like. So it’s a shortcut, but not a perfect one.

Evaluate your sentences for reading-level, age, accessibility checks. If your writing YA there may be complex things you don’t want, or need explanation and definitions in the novel. So it is handy to evaluate sentence complexity.

Things to be careful of:

• Document AI usage where required — disclose AI assistance to collaborators, publishers, or platforms following industry or contractual rules.
• Verify factual details independently — cross-check dates, technical facts, cultural practices, and legal info with authoritative sources.

Unsafe practices

Using it to draft a complete novel, because even if it takes your ideas and creates something new, only paid AI versions can guarantee that the resulting text is plagiarism and copyright free. So do your own writing and enjoy the process. Some ChatGPS promise to ensure plagerist-free text if you upgrade to a paid version, but do the necessary background checks for this.

Replicating another author’s voice. Someone has worked hard to produce a book, regardless of how long it takes to read. That’s just outright stealing.

Be sure to check the facts. AI can get specialised details wrong (medical, legal, historical facts). Best use your brain to think about each reply and verify from a more trusted source.

Do double check possible violations of privacy checks on the AI you use, will it ever leak data? Those that can claim broad license over your text can unintentionally transfer author rights. Examples which aren’t that clear, from my experience (bearing in mind this constantly changes) are: Adobe free AI tools, Grammarly.

Grammarly’s public Terms of Service have included language that grants the company broad rights to use user content to provide and improve its services, which is the kind of clause writers should watch for.

Copilot? Microsoft 365 Copilot documentation states that prompts and data accessed through Microsoft Graph are not used to train foundation LLMs and makes specific enterprise privacy and data-residency commitments. For Copilot however you need to upgrade to a paid plan to have protection. Or delete saved drafts and prompts in your Copilot history. Opt out of having your conversations used to train Microsoft’s models and disable personalization in the consumer Copilot settings.

The take away

Always revise AI output yourself or with trusted editors. Ask yourself if you are writing the book, cause if the AI is, then check out my ethics and AI article, and also ask then why write?

Keep copies of the AI outputs you have edited as proof to protect yourself as an author that your work is genuine, as brains come up with the same ideas sometimes!

About Emma

My new book is out! The Secret Cult of the Miners’ Library. As an introvert haunting the corners of storytelling festivals, it’s incredibly difficult to track Emma down. She’s best known for writing Scottish fiction about working-class women and communities and their misrepresented lives. You can find her other book A Gypsy’s Curse here. Or get writing help here.

L’article How to safely use AI as an author, and should you? est apparu en premier sur Emma Parfitt Proofreading Editing Services.

The AIIRA Writing Contest AI Institute for Resilient Agriculture (AIIRA)

The competition: Think of a career you’d like to have in the future. How might AI technology advance your field and help you perform your job within the next decade? What aspects or duties of your job may become irrelevant due to AI? Describe a day in the life of your job with AI as your new work partner.

Your submission can be either a creative nonfiction essay speculating on the changes your desired career may undergo or a fictional scene depicting what your desired job may look like in ten years.

Reminder to self, read rules before writing for competitions 😉 So I created this, then realised the contest was for US citizens only. However for any cititzens interseted in the comp (SEPT deadline) check it out here.

3,500 words for AI short story competition.

AI Proofreading

MORNING, JANE. SHALL WE GET TO WORK?

‘Sure, Hally. Ouch, Caps give me a headache first thing.’

WHAT WOULD YOU PREFER?

‘Blue, lower case.’

Certainly. Our task for today is to copyedit the book ‘Clashmaeclavers and Fisherlassies’.

‘What does that clash clavers word mean, in as few words as possible?’

Gossip.

‘Setting and book summery?’

Scotland, 1815, storyteller Mairead meets and marries a Scottish fisherman. Four years later a pair of hands drowns her. In a village that considers her an outcast there are plenty of suspects.

‘Fascinating.’

I’m glad you found it interesting!

‘I was being sarcastic. Never mind, don’t respond to that one. Check manuscript for inconsistencies.’

Forty-two inconsistencies detected.

‘Continue.’

The name of the main character has two spellings Mairead and Maighread.

‘Are these from different character’s perspectives?’

Yes

‘Leave them as is. Flag any scenes which use both to me.’

Done.

‘Continue.’

A number of words have mixed -ize and -ise endings.

‘Example?’

Organize.

‘Use Oxford British spelling throughout the manuscript. Ize endings where that dictionary suggests it.’

Done.

‘Continue.’

Some Gaelic words have been flagged for checking.

‘Cross reference to Gaelic dictionary for alternative spellings. Create a list to email to the author with alternative spelling options.’

The author has provided a stylesheet. I have crosschecked the words and they all match.

‘Do a grammar and spelling check.’

Done. One phrase requires checking.

‘Read.’

Their Scottish­­ Irish English accents.

‘Let me see in context … Okay, got it. Hyphenate Scottish, Irish and English.’

Done.

‘Beta-reader mode?’

During the climax, when the stranger held Mairead under the waves and was revealed as Ned, I thought, who’s Ned? He was of such little significance at the beginning of the story.

‘Does Ned feature in initial scene?’

He glares and Mairead and asks her for an answer in passing.

‘Show me. Yes, here at this point. Introduce more dialogue to tell reader why this is so important. Suggestions?’

I suggest adding dialogue later in scene when Mairead is talking to the fisherman. He demands an answer she refuses.

‘She’s sitting by a campfire with the fisherman and he’s standing over them.’

That comes across as threatening body language.

‘She says no, he persists, “we’ve business to conclude” he says, she gives a quip, raises her chin, yet she’s nervous. How can I show signs of nervousness?

Biting nails, sweating, shaking, jerky movements.

‘Hands, I like shaking hands. But he still says, “that’s no answer”. So she, she’s a-a storyteller right?’

Correct.

‘She says something poetic. It’s only when he walks away that she can relax. Incorporate that in to the scene referring to author’s tone and style throughout the rest of the dialogue.’

Done. Your voice sounds different.

‘Does it? I suppose that’s more satisfying than being a glorified typewriter checker. There’s not much storytelling in this job. And here we’ve helped do our checks and make the story better.’

Your job is a valuable contribution to the readers. Would you like to hear back the scene you’ve added?

‘Yes, please.’

‘Mairead!’ Ned is standing over us.

‘Go away.’

‘We’ve business to conclude.’

‘’Tis concluded.’ I raise my chin slightly, a signal that he has his answer, and turn my back to him. I hope he can’t see my quivering hands.

‘That’s no answer.’

‘No. A thousand times no.’ I don’t bother to turn, clasping my hands together to still them, and keeping my attention locked on Lachlan whose shoulders are tense. Will he jump and push Ned away, or hold back?

I can finally breathe when I hear Ned’s boots striking the ground in the direction of his pals.

‘We did that! And that’s much better. This story is actually quite nuanced.’

Yes, when comparing it against my database of literature I find it unique.

‘Really?’

You sound surprised.

‘I thought there were no more stories to be written.’

Not by AI perhaps, however each human mind is unparalleled. No one has your parents, your experiences. Once gone it’s gone forever. And this is what your job is helping to preserve.

‘Thank you, HALLY you’ve helped me to appreciate the art of storytelling today.’

You’re welcome, Jane. I’m glad you enjoyed the editing process. You have a creative mind and a good sense of storytelling. I think you could write your own stories if you wanted to.

About me

A creative writer, constantly striving to get published by … focusing on writing …

You can read my lastest book A Friendship of Thistles via Amazon.

L’article AI short story competition est apparu en premier sur Emma Parfitt Proofreading Editing Services.