Welcome back to What’s the Score?, the series in which we take novels from famous and bestselling authors, and put them to the test using the AutoCrit editing platform.
Report by report, we’ll give you an unparalleled look inside the workings of some of the most respected authors in every corner of the publishing industry.
Plus, you’ll get to see why the AutoCrit method is so powerful, and discover how to interpret report results for the maximum benefit of your manuscripts!
In this edition, we’re taking a trip to the furthest reaches of award-winning 1989 sci-fi novel Hyperion by Dan Simmons.
Please note this article contains short excerpts from the book, which you may want to avoid looking at too closely if you wish to avoid (minor) story spoilers.
DISCLAIMER: Note that this series is not designed to provide any kind of qualitative judgment nor a critique of the works dissected. As an exercise in exploration, What’s the Score? offers unique insight for us word nerds as to the linguistic makeup of specific novels, and evokes discussion of possible editing strategies in retrospect. All novels chosen for inspection have already proved immensely successful and stand as a testament to the author’s talents.
Summary Score and Fingerprint
Using AutoCrit’s “General Fiction” comparison setting, Hyperion delivers an impressive 85.2 summary score – a strong result for sure!
(You can compare all the scores we’ve uncovered so far by clicking here to see the rest of the series on this blog.)
Moving to the fingerprint, we can see the highest number of indicators appear in the Strong Writing category, followed by Repetition and then Word Choice. So let’s start breaking it down…
Looking at the dialogue tags included throughout Hyperion, we can see that Simmons prefers to keep them restricted to said and asked where possible – but that doesn’t mean he keeps it as a hard and fast rule.
Checking the most frequently used tags, we can see that said is the most common, followed by asked, whispered, and cried.
It might be surprising to see that even though said and asked are the recommended tags for dialogue in fiction, AutoCrit detects both here as being overused (hence having them marked in red).
This isn’t necessarily a problem. It simply means the tags have been detected more times than you would usually see in a fiction novel of this length – but that’s nothing to get too concerned about. Some books are more heavy on dialogue than others.
There is, of course, always the opportunity to re-frame dialogue in such a way that tags aren’t used at all. That’s an ideal way to cut down on their use (and avoid the possible pitfalls of using alternative tags).
Let’s check inside the manuscript to see AutoCrit’s numbered recommendations, and the individual words with which we’re dealing…
In the sidebar, we can see that recommended removal numbers are quite low, which is great – for all except the word whispered. With a recommended removal of about 35, that means more than half the total instances of this tag could be due for a trim.
While checking through the manuscript, we came across a spot where the tag gasped is used twice in quick succession and therefore stands out quite easily. With just 9 recommended removals for this word, this is a quick and easy example of a beneficial edit.
Still, from a high-level perspective, it wouldn’t take much work at all to bring Hyperion‘s dialogue tags into line with AutoCrit’s recommendations. It’s in great shape!
Next, let’s see what we’re dealing with in terms of adverbs attached to dialogue tags…
Just 3% of dialogue tags have an adverb attached, and that’s a seriously impressive result for what looks to be a pretty dialogue-heavy novel.
Leading the way for recommended trims are softly, formally, and truthfully. Just three words with a minimal number of recommended removals, too. Fantastic!
When it comes to repetition, our word clouds look pretty well spread out, here. In terms of single words, character names lead the way (which isn’t uncommon), and with phrases, the most common are either names or story-specific terms.
This is exactly what you’d expect to see, so nothing leaps out here as of particular concern. Where the real meat lies, however, is inside AutoCrit’s repetition reports. We spend some time jumping through repeated words, and came across one example that should immediately catch your eye…
With seven uses of the word passengers in just three paragraphs, this section really does stand out thanks to AutoCrit’s helpful highlights.
Admittedly, the section does read just fine, with perhaps only the third use of passengers feeling less than perfect because of the repetition. Swapping that one for an alternative would help add a spot of welcome variation.
Checking for repetition can be a lengthy effort when you’re dealing with a full manuscript, but it’s definitely worth it in terms of adding some extra “music” to your phrasing. If it feels daunting to begin with, try changing your AutoCrit settings to edit chapter by chapter, and tackle it on a per-chapter basis instead.
Strong Writing is our highest category for indicators in Hyperion, so let’s get started with adverbs. Thankfully, the count here is impressive – landing right between good and great on our gauge!
The author’s most frequently used adverbs aren’t detected as overused, but we should take a look at slightly, merely, quickly, softly, and a few others.
Removal numbers for the detected adverbs feel quite high compared to their overall use, here. For example, slightly sees 23 of 41 recommended for trimming, while merely sees around 30 of a total 39 suggested for removal. There may be some work to do, here!
In our sample screenshot above, you can see one paragraph containing five instances of an adverb – four of which are actually repetitious. In a case such as this, you might wish to excise most of these adverbs, perhaps leaving the first in place, or replace them with more metaphorical turns of phrase.
Next up, let’s check out Hyperion‘s filler words content…
Landing just above average, this is a pleasing result. Very, very few filler words appear to be overused – only even and seemed draw attention.
And while these two words are on our hit-list with this edit, removal numbers are low in comparison to total use. This looks like an easy quick win for Hyperion!
Let’s move on to our next target, Showing vs. Telling.
Once again, we’re just a little above average, which is actually not bad at all given the context of AutoCrit’s recommendations. All comparisons are made using the content of real best-selling novels, so even if you fall around the average mark, you know you’re in great company.
In terms of overuse, we’re looking at know, knew, felt, get, it’s, and a few other words. Showing vs. Telling is a very conceptual and fluid aspect of your book to tackle, which means every detected indicator isn’t guaranteed to need your attention – but it’s highly beneficial to take the time to scope them out and really dig into the heart of your writing.
(Side note: If you’re unsure how the word it could possibly indicate a lack of showing, check out our article Why “It” Matters in Your Book by clicking here.)
Recommended removal numbers have quite a range here, from a mere single instance of watch to 134 instances of there.
Reading through a portion of the manuscript with highlights intact, we couldn’t locate portions where we would recommend rewrites with showing vs. telling in mind (without being unreasonably bullish), but there are definitely opportunities to trim some instances of there, which can appear more as an unnecessary filler word in places.
Overall, in this case we’re quite happy to leave Hyperion sitting around the average mark. Improvements would be more easily found elsewhere, as the writing is enthralling enough as it is from an immersion perspective.
To finish this category, we have two more excellent results with just 1.29% of sentences containing redundant information (easily excusable for potential false positives) and 4.53% containing a known cliché phrase.
When it comes to clichés, remember that sometimes there’s no way around using a cliché turn of phrase if it’s within the correct context. If you use the term “on the table” in a metaphorical sense, you’d be falling foul of clichéd language – but if you use it in the literal sense when a character places something on the table, then that’s perfectly fine.
Either way, AutoCrit will detect it. Whether to change anything is a decision for you to make. This is your story, so just like we do, don’t be afraid to ignore or reject recommendations if they don’t fit with your goals.
Hyperion‘s winning streak continues as we check out the use of generic descriptions throughout the manuscript. Once again, we’re in “average” territory – coming in slightly below the dividing line and inside our positive blue area.
Our most frequently-used generic indicators are looked, great, low, large and wide – all of which should be quite obvious in their generic nature!
Hopping into the manuscript, we can see our recommended removal numbers aren’t too daunting. The word looked is far and away the most common with 197 uses throughout the novel, and around 25 of those might be suitable for a trim. In one instance, we came across a line that uses looked to frame what appears to be a “tell.” Would it be more beneficial to paint Father Hoyt’s appearance with a little more physical description? How exactly is he showing indications of indigestion that another human might pick up on?
Spend some time with this report and you can usually find at least a handful of instances where you might be brushing your way through description a little too broadly. As before, though, we’re happy to give Hyperion a solid thumbs up here with a positive result.
Moving on, and here’s a statistic that’s always of interest: the number of sentences throughout the novel that begin with either a pronoun or character name. Hyperion comes home with 42.57% of sentences beginning this way, which falls in line with the 40%-50% we’ve come to expect in fiction.
Sentences beginning with a “sentence starter” – usually a conjunction – form a mere 7.94% of the book.
With a quick check of our Point of View and Tense Consistency reports, it’s clear this is a novel written in the third person, past tense.
You’ll almost always see a healthy dose of present tense indicators showing up due to dialogue, but Hyperion does have a larger than expected chunk of first person narration detected. That’s something we’d definitely recommend checking on in your own work, just in case you’re shifting point of view too often, which can be confusing for readers.
In the case of Hyperion, it’s actually because of the book’s form, which is split into a Canterbury Tales-esque structure including personal journal entries and more traditional narrative.
So how does the emotional landscape of Hyperion come together through the author’s use of power words? Leading the way are fear-based words, backed up by encouragement/excitement, and a very similar smattering of the forbidden/secrets, lust and greed.
Put these together and you can certainly get an image of a tense space-faring adventure. This is a tale of intergalactic warfare and discovery, with a range of stories each with their own plot, but held under the same tonal banner.
Pacing & Momentum
When it comes to pacing, Hyperion offers a higher-than-expected detection rate of 10.81% slow-paced paragraphs. Given the book’s structure, however, this feels somewhat unavoidable.
With each tale, the author essentially has to restart the narrative arc, which affords much more leniency when viewing those red blocks at the bottom of the pacing chart. Indeed, to stop the book from feeling overly repetitive, some stories will naturally need to adopt an overall slower or faster pace based on their internal tone.
You might notice one particularly large spike near the end of the book. It’s a longer paragraph, but it doesn’t break the pace. Instead, it signals the arrival of a slower-paced section (note the large chunk of red shortly after the spike). This section is slower in comparison to what surrounds it, but the reading experience itself doesn’t suffer because of it. Simmons feels quite meticulous in his pacing, in fact, making Hyperion rarely less than involving.
Finally for this section, we can see an average of 12 words per sentence throughout Hyperion‘s tales, with 35 words per paragraph.
The Flesh Reading Ease score for Hyperion calculates at 73, meaning it’s considered fairly easy to read – conversational English with few complex words.
This goes to show that sci-fi, which can be notorious for its complex language, does not have to be inaccessible. Even the most imaginative ideas and scenarios can be communicated with clarity.
For more information on how Flesch Reading Ease is determined, and what you can do to make your writing more readable, take a look at the Flesch-Kincaid readability tests entry on Wikipedia and check out our Deep Dive Live video on the topic on YouTube.
Genre Comparison Score
To cap off our investigation, let’s compare Hyperion with AutoCrit’s Sci-Fi & Fantasy genre profile…
The result is a quick bump up to a summary score of 85.46. This may not be a huge shift, but we need to remember that this novel not only started at an excellent beginning point, but is particularly unusual in its structure compared to a “standard” novel. All things considered, even a slight increase is a great thing for Hyperion – a consistently strong result for this breakout work.
An eclectic author whose publishing history encompasses sci-fi, fantasy, horror, and the historical, Dan Simmons now has his very own comparison profile inside AutoCrit. Compare your writing style directly with his when you join AutoCrit Pro!
Receive all of the same recommendations you see in this article, plus more, all fine-tuned and adjusted so you can more closely match the word use and emotional power of the writers who inspired you to pick up the pen.
With new authors added on a continuous basis, you’re never short of a guiding hand that matches your intended audience. Join today for the full experience, or create your free account now to access some of our most popular interactive editing reports with no subscription.
Conquer The Five-Minute Agent Assessment That Can Make or Break Your Novel
Discover the three biggest manuscript "red flags" that send agents and publishers running for the hills...
And know exactly how to fix them – so you can escape the slush pile and make your stand on the shelves!
Don't get ignored. Get published. Enter your details to download your free guide now.