Adverbs *versus* Adverbs
====
TLDR: The word *real* is a modifier which like *very* works as an **intensifier**. Just like *very*, *real* can function not only as an adjective but also as a **special kind** of “adverb”.
As an adverb, it can only intensify other modifiers: it does **not** intensify verbs.
To compare *real* with *very*, consider how the very girl you were looking for might be very sweet-smelling, but you cannot *very look for her even if she is a real nice girl.
*Real* shares those properties, properties that are outside the stuffy eight-sided box of traditional part-of-speech tags some ‘grammarians’ believe we inherited from Latin.
----
*A rose by any other name. . . .*
---
First, here are some examples of sentences that discuss the fine fragrance of your sister’s rose:
1. Your sister’s rose smells real good. (***arguably** casual or informal*)
1. Your sister’s rose smells really good. (*unassailably fine)*
1. *Your sister’s rose real smells good. (***not grammatical***)
1. Your sister’s rose really smells good. (*again, a delightful aroma*)
1. Your sister’s rose really smells well. (*grammatical **but nonsensical***)
The lovely blossom does not work quite the way your lovely sister Rose does:
1. Your sister Rose smells real good. (***arguably** casual or informal*)
1. Your sister Rose smells really good. (*her fragrance is wonderful*)
1. *Your sister Rose real smells good. (***not grammatical***)
1. Your sister Rose really smells good. (*they might name a perfume after her*)<p>
1. Your sister Rose smells real well. (*arguably casual or informal*)
1. Your sister Rose smells really well. (*her olfactory powers are excellent*)
1. *Your sister Rose real smells well. (***not grammatical***)
1. Your sister Rose really smells well. (*this is about her olfactory powers again*)
----
Whatever you want to call *real*, it does not fit into neat and tidy categories. See also the question [What’s an adverb?](http://english.stackexchange.com/a/75510)
---
The *‑ly* and *‑like* derivational suffixes
===
The *‑ly* suffix is used to derive new modifiers from existing substantives (here read *nouns*), adjectives, or even at times verbs. That suffix has historically experienced a wide number of spellings, but the one that will be most familiar to English-speaking monoglots will be *‑like*.
If you happen to know some German, then the cognate *‑lich* will be a familiar suffix for you.
All Germanic languages work this way: not just German but also Gothic, Dutch, and the various Scandinavian tongues. Sometimes these are a bit tough to track down the origins of without a good dictionary. For example, words like *seem*, *seemly*, *unseemly* come from Old Norse.
Something that complicates this is that the OED recognizes **two different *‑ly* suffixes**.
1. one for making adjectives — call it *‑ly*<sup>1</sup>
2. the other for making adverbs — call it *‑ly*<sup>2</sup>
However, their history is somewhat muddled and clearly interrelated.
Indeed, it is not always clear whether the derivation is from a substantive or an adjective. As you yourself mention with the case of *ugly*, sometimes the original base form did not survive, or did so in another way which we no longer readily recognize. But it’s still there. (Oh, and *especial* is not non-existent; just a bit old-fashioned.)
Here are some examples of modifiers with *‑ly* at the end that may not leave a recognizable base in modern English.
* *Silly* is from *seely*, cognate to German *selig* and Dutch *zalig*.
* *Only* is from *one-ly*, and this is true whether used as an adjective or an adverb.
* *Especially* the adverb derives of course from the adjective *especial*, as in “an especial friend”. It is now more often just plain *special* in Modern English, but not always.
* *Ugly* comes from *ug-ly*, where *ug* comes from Old Norrs *ugg(r)* meaning fear or dread, and which in English survives only dialectally; *ug* is otherwise obsolete outside of dialect.
* Whether as adjective or adverb, *early* is from Old English *árlíce*, the positive degree of ON *ǽr* ‘ere’ + *-líce* ‘-ly’. At one point, people really did write *earlily* for the adverb corresponding to the adjective *early*, but no longer.
* Again irrespective of ad{jective,verb}, *daily* is from Old English *dæʒlíc*, a straightforward derivation of OE *dæʒ* with the *-líc* suffix. It is cognate to Old Norse *dagligr*.
* The adjective *timely* is clearly based on *time*, but does not occur in Old English and only rarely in Middle English. It does have an Old Norse cognate in *tímaligr* meaning temporal. The adverb version in Old English was *tímlíce*, from the OE substantive *tíma* time plus the *‑ly* suffix. So as an adverb it was certainly around in Old English, and that may be where the adjective version comes from.
I’ll get back to your original question RSN*, but first let’s look at both forms of the suffix.
<sub>Where *RSN* of course means **“real soon now”**. It does **not** mean “*really soon now”</sub>.
---
The OED on *‑ly*<sup>1</sup>
---
Here is a small excerpt from the OED for the first *‑ly* suffix:
> <p>The original Teut. adjs. in <i>-lîko-</i> were compounds of the
sb. <i>*lîkom</i> appearance, form, body (see <st>lich</st>). Thus
<i>*mannlîko-</i> (‘manly’) means etymologically ‘having the
appearance or form of a man’; <i>gôðolîko-</i> (‘goodly’) ‘having
a good appearance or form’, or ‘having the appearance or form of
what is good’. The primitive force of the suffix may therefore be
rendered by ‘having the appearance or form indicated by the first
element of the word’; but while in the historical Teut. langs. it
has remained capable of expressing this meaning, it has in all of
them acquired a much wider application.
<p>When appended to sbs., the most general senses of the suffix in
all Teut. langs. are ‘having the qualities appropriate to’,
‘characteristic of’, ‘befitting’. In English of all periods it has
been a prolific formative; the adjs. formed with it are most
frequently eulogistic, as in <i>kingly</i>, <i>knightly</i>,
<i>masterly</i>, <i>princely</i>, <i>queenly</i>, <i>scholarly</i>,
<i>soldierly</i> (cf. <st>manly</st>, <st>womanly</st> with
<i>mannish</i>, <i>womanish</i>); among the examples with dyslogistic
sense are <i>beastly</i>, <i>beggarly</i>, <i>cowardly</i>,
<i>dastardly</i>, <i>rascally</i>, <i>ruffianly</i>, <i>scoundrelly</i>.
In OE., as in other Teut. langs., the suffix had often the sense
‘of or pertaining to’; but the adjs. have, so far as this meaning
is concerned, been to a great extent superseded by synonyms of Latin
or Romanic etymology. Thus <i>manly</i> formerly admitted of the
senses now expressed by <i>human</i> and <i>masculine</i>; for one
of the older senses of <i>timely</i> we must now say <i>temporal</i>.
Another use of the suffix, common to English with other Teut. langs.,
is to form adjs. denoting periodic recurrence, as <i>daily</i>,
<i>hourly</i>, <i>monthly</i>, <i>nightly</i>, <i>weekly</i>,
<i>yearly</i>.
<p>When <i>-ly</i> is appended to an adj., the resulting derivative
adj. often connotes a quality related to or resembling that expressed
by its primary; cf., e.g., OE. <i>léof</i> ‘dear’ with <i>léoflic</i>
‘lovely’ (or, as it might be rendered, ‘such as becomes dear’). The
diminutive sense found in mod.G. <i>gelblich</i> yellowish,
<i>süsslich</i> sweetish, though a very easy development from the
original sense of the suffix, does not seem ever to have existed
in English. Even in OE. <i>-lic</i> had app. ceased to be used in
new formations from adjs.; the new adjs. f. adj. + <st>-ly</st>
that have arisen in ME. or in mod.E. seem to be from the advs.
There is a lot more where that came from, too.
The OED on *‑ly*<sup>2</sup>
---
For the second version, here is another brief excerpt:
> <p>The form-history of the suffix in Eng. is similar to that of <st>-ly</st><sup><b>1</b></sup>:
in ME. the OE. <i>-líce</i> was normally represented by <i>-līche</i>
(southern), <i>-līke</i> (northern), the compar. being <i>-lī̆ker</i>,
<i>-luker</i>, <i>-loker</i> (superl. <i>-est</i>).
<p>The form <i>-li</i>, <i>-ly</i>, which was current in East Midland
English in the 14th c., and became general in the 15th c., is
probably due to the influence of the ON. <i>-liga</i>. In the
strongly Scandinavianized dialect of the <i>Ormulum</i> (<i>c</i>
1200) <i>-liʒ</i> and <i>-like</i> are used indifferently, according
to the requirements of the metre. Where the positive ended in
<i>-li</i>, <i>-ly</i>, the comparative and superlative ended in
<i>-lier</i>, <i>-liest</i>. In the 15–17th c. forms like
<i>falslyer</i>, <i>traitorouslyer</i> (Malory), <i>softlier</i>,
<i>justlier</i>, <i>widelier</i> (Long <i>Barclay’s Argenis</i>
1625), <i>easilier</i>, <i>-est</i> (R. Baxter <i>Saving Faith</i>
1658) were common, but in later use the advs. in <i>-ly</i> are
compared with <i>more</i>, <i>most</i>, the inflexional forms being
only employed in poetry or for rhetorical effect.
<p>In OTeut. an adv. with this suffix must have implied the existence
of an adj. with the suffix corresponding to <st>-ly</st><sup><b>1</b></sup>. In
OE., however, there are several instances (e.g. <i>bealdlíce</i>
boldly, <i>swétlíce</i> sweetly) in which an adv. in <i>-líce</i>
has been formed directly from a simple adj. without the intervention
of an adj. in <i>-lic</i>. In ME. the number of these direct
formations was greatly increased, and when the final <i>-e</i>,
which was the original OE. adverb-making suffix, ceased to be
pronounced, it became usual to append <i>-ly</i> to an adj. as the
regular mode of forming an adv. of manner. It was, down to the 17th
c., somewhat frequently attached, with this function, even to adjs.
in <i>-ly</i>, as <i>earlily</i>, <i>godlily</i>, <i>kindlily</i>,
<i>livelily</i>, <i>lovelily</i>, <i>statelily</i>; but these
formations are now generally avoided as awkward, while on the other
hand it is felt to be ungraceful to use words like <i>godly</i>,
<i>goodly</i>, <i>lovely</i>, <i>mannerly</i>, <i>timely</i>, as
advs.; the difficulty is usually evaded by recourse to some
periphrastic form of expression. In examples belonging to the 16th
and 17th c. it is sometimes difficult to determine whether a writer
intended the adv. <i>goodly</i> to mean ‘in a good manner’ or ‘in
a goodly manner’, and there are other instances of similar ambiguity.
In the words denoting periodical recurrence, as <i>daily</i>,
<i>hourly</i>, the adj. and the adv. are now identical in form. A
solitary example of an adv. f. sb. + <st>-ly</st><sup><b>2</b></sup>
with no related adj. is <i>partly</i>. From the early part of the
16th c. the suffix has been added to ordinal numerals to form advs.
denoting serial position, as <i>firstly</i>, <i>secondly</i>,
<i>thirdly</i>, etc. (cf. F. <i>premièrement</i>, etc.).
<p>When <i>-ly</i> is attached to a disyllabic or polysyllabic adj.
in <i>-le</i>, the word is contracted, as in <i>ably</i>, <i>doubly</i>,
<i>singly</i>, <i>simply</i>; contractions of this kind occur already
in the 14th c., but examples of the uncontracted forms (e.g.
<i>doublely</i>) are found as late as the 17th c. <i>Whole</i> +
<st>-ly</st> becomes <i>wholly</i>, but in all other similar instances
the written <i>e</i> is retained before the suffix, e.g. in
<i>palely</i>, <i>vilely</i>, <i>puerilely</i>. Adjs. ending
graphically with <i>ll</i> lose one <i>l</i> before <i>-ly</i>, as
in <i>fully</i> (in southern Eng. commonly pronounced with a single
<i>l</i>, but in Scotland often with double or long <i>l</i>),
<i>dully</i> /ˈdʌllɪ/, <i>coolly</i> /ˈkuːllɪ/.
Adjs. of more than one syll. ending in <i>y</i> change <i>y</i> to
<i>i</i> before <i>-ly</i>, as in <i>merrily</i>; in formations
from monosyllabic adjs. the usage varies, e.g. <i>dryly</i>,
<i>drily</i>; <i>gayly</i>, <i>gaily</i> (cf. <st>daily</st>, which
is the only current form); <i>slyly</i>, <i>slily</i> (but always
<i>shyly</i>); <i>greyly</i>, <i>grayly</i> has always <i>y</i>.
Another orthographical point is the dropping of the <i>e</i> in the
two words <i>duly</i>, <i>truly</i>. It is unusual to append
<i>-ly</i> to an adj. in <i>-ic</i>; the ending of the adv. is
nearly always <st>-ically</st>,
even when the only current form of the adj. ends in <i>-ic</i>.
----
Back to the OP’s Question
====
When people say something like *real good*, they are using *real* as an adverb. This is permitted, especially in casual use in Australia and North America. The thing is, *real* as an adverb is not a general-purpose one. It only works on adjectives. The OED also notes that this is “Not common in standard use in southern England”. This may well be true; but there are many places where English is natively spoken outside of southern England.
About the adverb *real*, the OED says:
> real
> ====
> <p><b>B.</b> <i>adv.</i>
> ----
<b>1</b>. (Usually with adjs.) Really,
genuinely. Also more loosely in later use (orig. <i>Sc.</i> and
<i>U.S.</i>): Very, extremely. <small><blockquote>In early use
properly an adj. qualifying the phrase (‘good turn’, etc.) which
follows, and only at a later period apprehended as an adv. qualifying
the adj. (‘good’, etc.). Not common in standard use in southern
England except to some extent in the orig. construction.
</blockquote></small>
<p><ul><small> <li><st><b>1658</b></st> <i>Whole Duty Man</i> xiii.
§35 ― The reallest good turn that can be done from one man to
another. <li><st><b>1718</b></st> <st>J. Fox</st> <i>Wanderer</i>
No. 17. 116 ― An Opportunity of doing a real good Office.
<li><st><b>1771</b></st> <st>Mrs. Griffith</st> <i>Hist. Lady
Barton</i> II. 283 ― The burning of three real good and substantial
houses in this town. <li><st><b>1827</b></st> <st>R. H. Froude</st>
<i>Rem.</i> (1838) I. 448 ― Last Friday was a real fine day.
<li><st><b>1885</b></st> <st>G. Allen</st> <i>Babylon</i> vi, ― It
looks real nice. <li><st><b>1887</b></st> <st>Mabel Wetheral</st>
<i>Two N.-C. Maids</i> xxv. 174, ― I was real put out to think how
[etc.]. <li><st><b>1939</b></st> <i>War Illustr.</i> 28 Oct. 219/1
― If I had not been on fire I could easily have shot down two more.
It was real bad luck, but my pals accounted for three besides the
one I hit. <li><st><b>1943</b></st> <st>K. Tennant</st> <i>Ride
on Stranger</i> viii. 77 ― He’s real clever. <li><st><b>1959</b></st>
<st>J. Ludwig</st> in <i>Tamarack Rev.</i> Summer 7 ― Some day she’d
get real tough with her son Sidney. <li><st><b>1968</b></st>
<i>Globe & Mail</i> (Toronto) 17 Feb. 50/3 (Advt.), ― Austin
Healey Sprite black, radio, a real nice car. <li><st><b>1968</b></st>
<st>K. Weatherly</st> <i>Roo Shooter</i> 111 ― It was real heavy
going, and I must have dried the flamin’ plugs and points twenty
times. <li><st><b>1976</b></st> <i>Daily Mirror</i> 18 Mar. 24/4
― I’m havin’ a rest-I feel real listless.</small></ul>
<p><b>2</b>. (with advbs.) <i>colloq.</i> (chiefly <i>N. Amer.</i>
and <i>Austral.</i>).
<p><ul><small> <li><st><b>1893</b></st> <st>H. A. Shands</st> <i>Some
Peculiarities of Speech in Mississippi</i> 52 ― Real down... Used
by cultivated whites to mean <i>exceedingly</i> or <i>extremely</i>.
A thing that is extremely nice is said to be <i>real down</i> nice.
<li><st><b>1924</b></st> <st>J. C. French</st> <i>Writing</i> x.
290 ― Avoid: They live <i>good</i> in that camp (say live <i>well</i>),
I <i>sure</i> will write <i>real</i> soon (say <i>surely</i> will,
<i>really</i> soon). <li><st><b>1933</b></st> <st>R. L. Pooley</st>
in <i>Amer. Speech</i> VIII. 61/2 ― One such [grammarian], commenting
on the sentence, ‘I will write <i>real</i> soon,’ corrects <i>real</i>
to read <i>really</i>. This is utter nonsense. No one ever says
<i>I will write really soon</i>... It simply isn’t English.
<li><st><b>1942</b></st> <st>Z. N. Hurston</st> in A. Dundes <i>Mother
Wit</i> (1973) 225/1 ― De man looked at me real hard for dat.
<li><st><b>1947</b></st> <st>K. Tennant</st> <i>Lost Haven</i> xix.
317 ― Everyone said she was lucky... Everything fell out ‘real nice’
for her. <li><st><b>1959</b></st> <i>Weekly Times</i> (Melbourne)
30 Sept. (Advt.), ― How about picking up your phone and asking your
B.F.E. dealer to arrange a free demonstration of a ‘35’ on your
property real soon? <li><st><b>1967</b></st> <st>G. Jackson</st>
<i>Let.</i> 13 July in <i>Soledad Brother</i> (1971) 121, ― I felt
real bad about that. <li><st><b>1975</b></st> <st>D. Lodge</st>
<i>Changing Places</i> ii. 57 ― You and I must have lunch together
real soon.</small></ul>
This happens a great deal. If you [look for the word *real* in COCA followed by an adjective](http://corpus.byu.edu/coca/?c=coca&q=32400326), which is something that’s real easy to do, you will find that the top ten adjectives that follow *real* immediately are:
1. good 1378
2. quick 858
3. hard 541
4. bad 483
5. nice 422
6. big 341
7. fast 258
8. political 253
9. close 240
10. simple 222
Other popular choices include *easy* and *happy*.
So as I said, it happens — and not just a little, either: it happens a lot. The OED describes it, then [COCA](http://corpus.byu.edu/coca/) gives you a feel for the common adjectives which the adverb *real* collocates with.
Now, this situation can be uncannily knickersnitting to certain people. The kind of people it upsets are the ones more interested in telling you what you should say then they are in listening to what you do say. Prescriptive ‘grammarians’ and hen-pecked proofreaders who’ve been given their marching orders oft will auto-correct that *real good* to *really good*, at least in formal writing. but there is a long history of *real* being used without an *‑ly* suffix.
But it is how real people really talk, and so to the descriptivist, it merely represents an alternative way of saying the same thing.
However, it is now looked at rather poorly in some quarters, though. Indeed, its use has come to be a **class marker**. So do not be surprised to see it turned into *really* when used as an intensifier.
All these rigid “rules” about parts of speech are after-the-fact creations. There are more kinds of words than the original eight parts of speech, and intensifiers are one of them. Fast cars can go real fast, and they can go really fast, but they cannot speed very. Notice how a car that’s really speeding is different from a car this is speeding really.
Real language is much more complicated and nuanced than a third-grader’s grammar book will ever reveal. If you want a simplified version of reality, by all means, read the third-grader’s book. Just understand that it isn’t real good at telling you anything about how real people really talk. It’s all something something of an inconvenient lie, or at best, so simplified a version of reality as to no longer bear much semblance of the same.