14370824
14370824
com
OR CLICK HERE
DOWLOAD EBOOK
(Ebook) Python Natural Language Processing: Advanced machine learning and deep
learning techniques for natural language processing by Jalaj Thanaki ISBN
9781787121423, 1787121429
https://ebooknice.com/product/python-natural-language-processing-advanced-
machine-learning-and-deep-learning-techniques-for-natural-language-
processing-7212480
ebooknice.com
(Ebook) Deep Learning for Natural Language Processing: Develop Deep Learning Models
for Natural Language in Python by Jason Brownlee
https://ebooknice.com/product/deep-learning-for-natural-language-processing-
develop-deep-learning-models-for-natural-language-in-python-11575184
ebooknice.com
(Ebook) Deep Learning for Natural Language Processing: Develop Deep Learning Models
for your Natural Language Problems by Jason Brownlee
https://ebooknice.com/product/deep-learning-for-natural-language-processing-
develop-deep-learning-models-for-your-natural-language-problems-42525080
ebooknice.com
(Ebook) Natural Language Processing Recipes: Unlocking Text Data with Machine
Learning and Deep Learning Using Python by Akshay Kulkarni, Adarsha Shivananda ISBN
9781484273500, 1484273508
https://ebooknice.com/product/natural-language-processing-recipes-unlocking-
text-data-with-machine-learning-and-deep-learning-using-python-34204404
ebooknice.com
(Ebook) Natural Language Processing Recipes: Unlocking Text Data with Machine
Learning and Deep Learning Using Python by Akshay Kulkarni, Adarsha Shivananda ISBN
9781484273517, 9781484273500, 1484273508, 1484273516
https://ebooknice.com/product/natural-language-processing-recipes-unlocking-
text-data-with-machine-learning-and-deep-learning-using-python-34204734
ebooknice.com
(Ebook) Deep Learning for Natural Language Processing: Creating Neural Networks with
Python by Palash Goyal, Sumit Pandey, Karan Jain ISBN 9781484236840, 148423684X
https://ebooknice.com/product/deep-learning-for-natural-language-processing-
creating-neural-networks-with-python-50195740
ebooknice.com
(Ebook) Applied Reinforcement Learning with Python: With OpenAI Gym, Tensorflow, and
Keras by Beysolow II, Taweh ISBN 9781484251263, 1484251261
https://ebooknice.com/product/applied-reinforcement-learning-with-python-with-
openai-gym-tensorflow-and-keras-55666056
ebooknice.com
(Ebook) Getting started with Deep Learning for Natural Language Processing: Learn
how to build NLP applications with Deep Learning by Sunil Patel ISBN 9789389898118,
9389898110
https://ebooknice.com/product/getting-started-with-deep-learning-for-natural-
language-processing-learn-how-to-build-nlp-applications-with-deep-
learning-36407330
ebooknice.com
(Ebook) Deep Learning for Natural Language Processing (MEAP V07) by Stephan
Raaijmakers ISBN 9781617295447, 1617295442
https://ebooknice.com/product/deep-learning-for-natural-language-processing-
meap-v07-12156196
ebooknice.com
Applied Natural
Language Processing
with Python
Implementing Machine Learning
and Deep Learning Algorithms for
Natural Language Processing
—
Taweh Beysolow II
Applied Natural
Language Processing
with Python
Implementing Machine
Learning and Deep Learning
Algorithms for Natural
Language Processing
Taweh Beysolow II
Applied Natural Language Processing with Python
Taweh Beysolow II
San Francisco, California, USA
v
Table of Contents
vi
Table of Contents
Index�������������������������������������������������������������������������������������������������145
vii
About the Author
Taweh Beysolow II is a data scientist and
author currently based in San Francisco,
California. He has a bachelor’s degree in
economics from St. Johns University and a
master’s degree in applied statistics from
Fordham University. His professional
experience has included working at Booz
Allen Hamilton, as a consultant and in various
startups as a data scientist, specifically
focusing on machine learning. He has applied machine learning to federal
consulting, financial services, and agricultural sectors.
ix
About the Technical Reviewer
Santanu Pattanayak currently works at GE
Digital as a staff data scientist and is the author
of the deep learning book Pro Deep Learning
with TensorFlow: A Mathematical Approach
to Advanced Artificial Intelligence in Python
(Apress, 2017). He has more than eight years of
experience in the data analytics/data science
field and a background in development and
database technologies. Prior to joining GE,
Santanu worked at companies such as RBS,
Capgemini, and IBM. He graduated with a degree in electrical engineering
from Jadavpur University, Kolkata, and is an avid math enthusiast. Santanu
is currently pursuing a master’s degree in data science from the Indian
Institute of Technology (IIT), Hyderabad. He also devotes his time to data
science hackathons and Kaggle competitions, where he ranks within the
top 500 across the globe. Santanu was born and brought up in West Bengal,
India, and currently resides in Bangalore, India, with his wife.
xi
Acknowledgments
A special thanks to Santanu Pattanayak, Divya Modi, Celestin Suresh
John, and everyone at Apress for the wonderful experience. It has been a
pleasure to work with you all on this text. I couldn’t have asked for a better
team.
xiii
Visit https://ebooknice.com to
discover a wide range of
eBooks across various genres.
Enjoy exclusive deals and
discounts to enhance your
reading experience. Start your
digital reading journey today!
Introduction
Thank you for choosing Applied Natural Language Processing with Python
for your journey into natural language processing (NLP). Readers should
be aware that this text should not be considered a comprehensive study
of machine learning, deep learning, or computer programming. As such,
it is assumed that you are familiar with these techniques to some degree.
Regardless, a brief review of the concepts necessary to understand the
tasks that you will perform in the book is provided.
After the brief review, we begin by examining how to work with raw
text data, slowly working our way through how to present data to machine
learning and deep learning algorithms. After you are familiar with some
basic preprocessing algorithms, we will make our way into some of the
more advanced NLP tasks, such as training and working with trained
word embeddings, spell-check, text generation, and question-and-answer
generation.
All of the examples utilize the Python programming language and
popular deep learning and machine learning frameworks, such as scikit-
learn, Keras, and TensorFlow. Readers can feel free to access the source
code utilized in this book on the corresponding GitHub page and/or try
their own methods for solving the various problems tackled in this book
with the datasets provided.
xv
CHAPTER 1
What Is Natural
Language
Processing?
Deep learning and machine learning continues to proliferate throughout
various industries, and has revolutionized the topic that I wish to discuss
in this book: natural language processing (NLP). NLP is a subfield of
computer science that is focused on allowing computers to understand
language in a “natural” way, as humans do. Typically, this would refer to
tasks such as understanding the sentiment of text, speech recognition, and
generating responses to questions.
NLP has become a rapidly evolving field, and one whose applications
have represented a large portion of artificial intelligence (AI)
breakthroughs. Some examples of implementations using deep learning
are chatbots that handle customer service requests, auto-spellcheck on cell
phones, and AI assistants, such as Cortana and Siri, on smartphones. For
those who have experience in machine learning and deep learning, natural
language processing is one of the most exciting areas for individuals to
apply their skills. To provide context for broader discussions, however, let’s
discuss the development of natural language processing as a field.
2
Chapter 1 What Is Natural Language Processing?
The SLP model is seen to be in part due to Alan Turing’s research in the
late 1930s on computation, which inspired other scientists and researchers
to develop different concepts, such as formal language theory.
Moving forward to the second half of the twentieth century, NLP starts
to bifurcate into two distinct groups of thought: (1) those who support a
symbolic approach to language modelling, and (2) those who support a
stochastic approach. The former group was populated largely by linguists
who used simple algorithms to solve NLP problems, often utilizing pattern
recognition. The latter group was primarily composed of statisticians
and electrical engineers. Among the many approaches that were popular
with the second group was Bayesian statistics. As the twentieth century
progressed, NLP broadened as a field, including natural language
understanding (NLU) to the problem space (allowing computers to react
accurately to commands). For example, if someone spoke to a chatbot and
asked it to “find food near me,” the chatbot would use NLU to translate this
sentence into tangible actions to yield a desirable outcome.
Skip closer to the present day, and we find that NLP has experienced
a surge of interest alongside machine learning’s explosion in usage over
the past 20 years. Part of this is due to the fact that large repositories of
labeled data sets have become more available, in addition to an increase in
computing power. This increase in computing power is largely attributed
to the development of GPUs; nonetheless, it has proven vital to AI’s
development as a field. Accordingly, demand for materials to instruct
data scientists and engineers on how to utilize various AI algorithms has
increased, in part the reason for this book.
Now that you are aware of the history of NLP as it relates to the present
day, I will give a brief overview of what you should expect to learn. The
focus, however, is primarily to discuss how deep learning has impacted
NLP, and how to utilize deep learning and machine learning techniques to
solve NLP problems.
3
Chapter 1 What Is Natural Language Processing?
TensorFlow
One of the groundbreaking releases in open source software, in addition
to machine learning at large, has undoubtedly been Google’s TensorFlow.
It is an open source library for deep learning that is a successor to Theano,
a similar machine learning library. Both utilize data flow graphs for
4
Chapter 1 What Is Natural Language Processing?
targets
5
Chapter 1 What Is Natural Language Processing?
'output': tf.Variable(tf.random_normal([state_size,
n_classes]))}
biases = {'input': tf.Variable(tf.random_normal([1, state_
size])),
'output': tf.Variable(tf.random_normal([1, n_classes]))}
Keras
Due to the slow development process of applications in TensorFlow,
Theano, and similar deep learning frameworks, Keras was developed for
prototyping applications, but it is also utilized in production engineering
for various problems. It is a wrapper for TensorFlow, Theano, MXNet, and
DeepLearning4j. Unlike these frameworks, defining a computational graph
is relatively easy, as shown in the following Keras demo code.
def create_model():
model = Sequential()
model.add(ConvLSTM2D(filters=40, kernel_size=(3, 3),
input_shape=(None, 40, 40, 1),
padding='same', return_sequences=True))
model.add(BatchNormalization())
7
Chapter 1 What Is Natural Language Processing?
activation='sigmoid',
padding='same', data_format='channels_last'))
model.compile(loss='binary_crossentropy', optimizer='adadelta')
return model
Although having the added benefit of ease of use and speed with
respect to implementing solutions, Keras has relative drawbacks when
compared to TensorFlow. The broadest explanation is that Keras
users have considerably less control over their computational graph
than TensorFlow users. You work within the confines of a sandbox
when using Keras. TensorFlow is better at natively supporting more
complex operations, and providing access to the most cutting-edge
implementations of various algorithms.
Theano
Although it is not covered in this book, it is important in the progression
of deep learning to discuss Theano. The library is similar to TensorFlow
in that it provides developers with various computational functions (add,
matrix multiplication, subtract, etc.) that are embedded in tensors when
building deep learning and machine learning models. For example, the
following is a sample Theano code.
8
Chapter 1 What Is Natural Language Processing?
if __name__ == '__main__':
model_predict()
9
Visit https://ebooknice.com to
discover a wide range of
eBooks across various genres.
Enjoy exclusive deals and
discounts to enhance your
reading experience. Start your
digital reading journey today!
Chapter 1 What Is Natural Language Processing?
T opic Modeling
In Chapter 4, we discuss more advanced uses of deep learning, machine
learning, and NLP. We start with topic modeling and how to perform it via
latent Dirichlet allocation, as well as non-negative matrix factorization.
Topic modeling is simply the process of extracting topics from documents.
You can use these topics for exploratory purposes via data visualization or
as a preprocessing step when labeling data.
W
ord Embeddings
Word embeddings are a collection of models/techniques for mapping
words (or phrases) to vector space, such that they appear in a high-
dimensional field. From this, you can determine the degree of similarity,
or dissimilarity, between one word (or phrase, or document) and another.
When we project the word vectors into a high-dimensional space, we can
envision that it appears as something like what’s shown in Figure 1-3.
10
Chapter 1 What Is Natural Language Processing?
walked
swam
walking
swimming
Verb tense
Figure 1-3. Visualization of word embeddings
11
Chapter 1 What Is Natural Language Processing?
Summary
The purpose of this book is to familiarize you with the field of natural
language processing and then progress to examples in which you
can apply this knowledge. This book covers machine learning where
necessary, although it is assumed that you have already used machine
learning models in a practical setting prior.
While this book is not intended to be exhaustive nor overly academic,
it is my intention to sufficiently cover the material such that readers are
able to process more advanced texts more easily than prior to reading
it. For those who are more interested in the tangible applications of NLP
as the field stands today, it is the vast majority of what is discussed and
shown in examples. Without further ado, let’s begin our review of machine
learning, specifically as it relates to the models used in this book.
12
CHAPTER 2
Review of Deep
Learning
You should be aware that we use deep learning and machine learning
methods throughout this chapter. Although the chapter does not provide
a comprehensive review of ML/DL, it is critical to discuss a few neural
network models because we will be applying them later. This chapter also
briefly familiarizes you with TensorFlow, which is one of the frameworks
utilized during the course of the book. All examples in this chapter use toy
numerical data sets, as it would be difficult to both review neural networks
and learn to work with text data at the same time.
Again, the purpose of these toy problems is to focus on learning how
to create a TensorFlow model, not to create a deployable solution. Moving
forward from this chapter, all examples focus on these models with text data.
non-linear, rendering the SLP null and void. MLPs are able to overcome
this shortcoming—specifically because MLPs have multiple layers. We’ll
go over this detail and more in depth while walking through some code to
make the example more intuitive. However, let’s begin by looking at the
MLP visualization shown in Figure 2-1.
14
Chapter 2 Review of Deep Learning
This Python function contains all the TensorFlow code that forms the
body of the neural network. In addition to defining the graph, this function
invokes the TensorFlow session that trains the network and makes
predictions. We’ll begin by walking through the function, line by line, while
tying the code back to the theory behind the model.
First, let’s address the arguments in our function: train_data is the
variable that contains our training data; in this example; it is the returns of
specific stocks over a given period of time. The following is the header of
our data set:
15
Chapter 2 Review of Deep Learning
1 N
( )
2
Si =1 hq ( x ) - y i
i
q t +1 = q t - h (2.1)
N
q t +1 = q t - h
1
N
( i
)
Si =1 2 hq ( x ) - y i Ñq hq ( x )
i
16
Chapter 2 Review of Deep Learning
Each unit in a neural network (with the exception of the input layer)
receives the weighted sum of inputs multiplied by weights, all of which are
summed with a bias. Mathematically, this can be described in Equation 2.2.
y = f ( x ,w T ) + b (2.2)
In neural networks, the parameters are the weights and biases. When
referring to Figure 2-1, the weights are the lines that connect the units in
a layer to one another and are typically initialized by randomly sampling
from a normal distribution. The following is the code where this occurs:
weights = {'input': tf.Variable(tf.random_normal([train_x.
shape[1], num_hidden])),
'hidden1': tf.Variable(tf.random_normal([num_
hidden, num_hidden])),
'output': tf.Variable(tf.random_normal([num_hidden,
1]))}
biases = {'input': tf.Variable(tf.random_normal([num_
hidden])),
'hidden1': tf.Variable(tf.random_normal([num_
hidden])),
'output': tf.Variable(tf.random_normal([1]))}
Because they are part of the computational graph, weights and biases
in TensorFlow must be initialized as TensorFlow variables with the tf.
Variable(). TensorFlow thankfully has a function that generates numbers
randomly from a normal distribution called tf.random_normal(), which
takes an array as an argument that specifies the shape of the matrix that you
are creating. For people who are new to creating neural networks, choosing
the proper dimensions for the weight and bias units is a typical source of
frustration. The following are some quick pointers to keep in mind :
17
Chapter 2 Review of Deep Learning
The more general problem associated with weights that are initialized
at the same location is that it makes the network susceptible to getting
stuck in local minima. Let’s imagine an error function, such as the one
shown in Figure 2-2.
18
Random documents with unrelated
content Scribd suggests to you:
men as must have a faith to inspire them. And that is so. Assaulted
by reason, by logic, argument, philanthropy, progress directed
against his peculiar institutions, the Southerner at last is driven to a
fanaticism—a sacred faith which is above all reason or logical attack
in the propriety, righteousness, and divinity of slavery.
The chaplain, a venerable old man, loudly invoked curses on the
heads of the enemy, and blessings on the arms and councils of the
New State. When he was done, Mr. Howell Cobb, a fat, double-
chinned, mellow-eyed man, rapped with his hammer on the desk
before the chair on which he sat as speaker of the assembly, and the
house proceeded to business. I could fancy that, in all but garments,
they were like the men who first conceived the great rebellion which
led to the independence of this wonderful country—so earnest, so
grave, so sober, and so vindictive—at least, so embittered against
the power which they consider tyrannical and insulting.
The word “liberty” was used repeatedly in the short time allotted
to the public transaction of business and the reading of documents;
the Congress was anxious to get to its work, and Mr. Howell Cobb
again thumped his desk and announced that the house was going
into “secret session,” which intimated that all persons who were not
members should leave. I was introduced to what is called the floor
of the house, and had a delegate’s chair, and of course I moved
away with the others, and with the disappointed ladies and men
from the galleries, but one of the members, Mr. Rhett, I believe, said
jokingly: “I think you ought to retain your seat. If the ‘Times’ will
support the South, we’ll accept you as a delegate.” I replied that I
was afraid I could not act as a delegate to a Congress of Slave
States. And, indeed, I had been much affected at the slave auction
held just outside the hotel, on the steps of the public fountain, which
I had witnessed on my way to the capitol. The auctioneer, who was
an ill-favoured, dissipated-looking rascal, had his “article” beside him
on, not in, a deal packing-case—a stout young negro badly dressed
and ill-shod, who stood with all his goods fastened in a small bundle
in his hand, looking out at the small and listless gathering of men,
who, whittling and chewing, had moved out from the shady side of
the street as they saw the man put up. The chattel character of
slavery in the States renders it most repulsive. What a pity the
nigger is not polypoid—so that he could be cut up in chunks, and
each chunk should reproduce itself!
A man in a cart, some volunteers in coarse uniforms, a few Irish
labourers in a long van, and four or five men in the usual black coat,
satin waistcoat, and black hat, constituted the audience, whom the
auctioneer addressed volubly: “A prime field-hand! Just look at him
—good-natered, well-tempered; no marks, nary sign of bad about
him! En-i-ne hunthered—only nine hun-ther-ed and fifty dol’rs for
’em! Why, it’s quite rad-aklous! Nine hundred and fifty dol’rs! I can’t
raly——That’s good. Thank you, sir. Twenty-five bid—nine hun-therd
and seventy-five dol’rs for this most useful hand.” The price rose to
one thousand dollars, at which the useful hand was knocked down
to one of the black hats near me. The auctioneer and the negro and
his buyer all walked off together to settle the transaction, and the
crowd moved away.
“That nigger went cheap,” said one of them to a companion, as he
walked towards the shade. “Yes, Sirr! Niggers is cheap now—that’s a
fact.” I must admit that I felt myself indulging in a sort of reflection
whether it would not be nice to own a man as absolutely as one
might possess a horse—to hold him subject to my will and pleasure,
as if he were a brute beast without the power of kicking or biting—to
make him work for me—to hold his fate in my hands: but the
thought was for a moment. It was followed by disgust.
I have seen slave markets in the East, where the traditions of the
race, the condition of family and social relations divest slavery of the
most odious characteristics which pertain to it in the States; but the
use of the English tongue in such a transaction, and the idea of its
taking place among a civilised Christian people, produced in me a
feeling of inexpressible loathing and indignation. Yesterday I was
much struck by the intelligence, activity, and desire to please of a
good-looking coloured waiter, who seemed so light-hearted and
light-coloured I could not imagine he was a slave. So one of our
party, who was an American, asked him: “What are you, boy—a free
nigger?” Of course he knew that in Alabama it was most unlikely he
could reply in the affirmative. The young man’s smile died away from
his lips, a flush of blood embrowned the face for a moment, and he
answered in a sad, low tone: “No, sir! I b’long to Massa Jackson,”
and left the room at once. As I stood at an upper window of the
capitol, and looked on the wide expanse of richly-wooded, well-
cultivated land which sweeps round the hill-side away to the horizon,
I could not help thinking of the misery and cruelty which must have
been borne in tilling the land and raising the houses and streets of
the dominant race before whom one nationality of coloured people
has perished within the memory of man. The misery and cruelty of
the system are established by the advertisements for runaway
negroes, and by the description of the stigmata on their persons—
whippings and brandings, scars and cuts—though these, indeed, are
less frequent here than in the border States.
On my return, the Hon. W. M. Browne, Assistant-Secretary of
State, came to visit me—a cadet of an Irish family, who came to
America some years ago, and having lost his money in land
speculations, turned his pen to good account as a journalist, and
gained Mr. Buchanan’s patronage and support as a newspaper editor
in Washington. There he became intimate with the Southern
gentlemen, with whom he naturally associated in preference to the
Northern members; and when they went out, he walked over along
with them. He told me the Government had already received
numerous—I think he said 400—letters from shipowners applying for
letters of marque and reprisal. Many of these applications were from
merchants in Boston, and other maritime cities in the New England
States. He further stated that the President was determined to take
the whole control of the army, and the appointments to command in
all ranks of officers into his own hands.
There is now no possible chance of preserving the peace or of
averting the horrors of war from these great and prosperous
communities. The Southern people, right or wrong, are bent on
independence and on separation, and they will fight to the last for
their object.
The press is fanning the flame on both sides: it would be difficult
to say whether it or the telegraphs circulate lies most largely; but
that as the papers print the telegrams they must have the palm. The
Southerners are told there is a reign of terror in New York—that the
7th New York Regiment has been captured by the Baltimore people
—that Abe Lincoln is always drunk—that General Lee has seized
Arlington heights, and is bombarding Washington. The New York
people are regaled with similar stories from the South. The
coincidence between the date of the skirmish at Lexington and of
the attack on the 6th Massachusetts Regiment at Baltimore is not so
remarkable as the fact, that the first man who was killed at the latter
place, 86 years ago, was a direct descendant of the first of the
colonists who was killed by the royal soldiery. Baltimore may do the
same for the South which Lexington did for all the Colonies. Head-
shaving, forcible deportations, tarring and feathering are
recommended and adopted as specifics to produce conversion from
erroneous opinions. The President of the United States has called
into service of the Federal Government 42,000 volunteers, and
increased the regular army by 22,000 men, and the navy by 18,000
men. If the South secede, they ought certainly to take over with
them some Yankee hotel keepers. This “Exchange” is in a frightful
state—nothing but noise, dirt, drinking, wrangling.
CHAPTER XXIII.
The vessel was nothing more than a vast wooden house, of three
separate storeys, floating on a pontoon which upheld the engine,
with a dining-hall or saloon on the second storey surrounded by
sleeping-berths, and a nest of smaller rooms up-stairs; on the metal
roof was a “musical” instrument called a “calliope,” played like a
piano by keys, which acted on levers and valves, admitting steam
into metal cups, where it produced the requisite notes—high,
resonant, and not unpleasing at a moderate distance. It is 417 miles
to Mobile, but at this season the steamer can maintain a good rate
of speed as there is very little cotton or cargo to be taken on board
at the landings, and the stream is full.
The river is about 200 yards broad, and of the colour of chocolate
and milk, with high, steep, wooded banks, rising so much above the
surface of the stream, that a person on the upper deck of the
towering Southern Republic, cannot get a glimpse of the fields and
country beyond. High banks and bluffs spring up to the height of
150 or even 200 feet above the river, the breadth of which is so
uniform as give the Alabama the appearance of a canal, only
relieved by sudden bends and rapid curves. The surface is covered
with masses of drift-wood, whole trees, and small islands of
branches. Now and then a sharp, black, fang-like projection standing
stiffly in the current gives warning of a snag, but the helmsman, who
commands the whole course of the river, from an elevated house
amidships on the upper deck, can see these in time; and at night
pine boughs are lighted in iron cressets at the bows to illuminate the
water.
The captain, who was not particular whether his name was spelt
Maher, or Meaher, or Meagher (les trois se disent), was evidently a
character—perhaps a good one. One with a grey eye full of cunning
and of some humour, strongly marked features, and a very Celtic
mouth of the Kerry type. He soon attached himself to me, and
favoured me with some wonderful yarns, which I hope he was not
foolish enough to think I believed. One relating to a wholesale
destruction and massacre of Indians, he narrated with evident gusto.
Pointing to one of the bluffs, he said that some thirty years ago the
whole of the Indians in the district being surrounded by the whites,
betook themselves to that spot and remained there without any
means of escape, till they were quite starved out. So they sent down
to know if the whites would let them go, and it was agreed that they
should be permitted to move down the river in boats. When the day
came, and they were all afloat, the whites anticipated the boat-
massacre of Nana Sahib at Cawnpore, and destroyed the helpless
red skins. Many hundreds thus perished, and the whole affair was
very much approved of.
The value of land on the sides of this river is great, as it yields
nine to eleven bales of cotton to the acre—worth 10l. a bale at
present prices. The only evidences of this wealth to be seen by us
consisted of the cotton sheds on the top of the banks, and slides of
timber, with steps at each side down to the landings, so constructed
that the cotton bales could be shot down on board the vessel. These
shoots and staircases are generally protected by a roof of planks,
and lead to unknown regions inhabited by niggers and their masters,
the latter all talking politics. They never will, never can be conquered
—nothing on earth could induce them to go back into the Union.
They will burn every bale of cotton, and fire every house, and lay
waste every field and homestead before they will yield to the
Yankees. And so they talk through the glimmering of bad cigars for
hours.
The management of the boat is dexterous,—as she approaches a
landing place, the helm is put hard over, to the screaming of the
steam-pipe and the wild strains of “Dixie” floating out of the throats
of the calliope, and as the engines are detached, one wheel is
worked forward, and the other backs water, so she soon turns head
up stream, and is then gently paddled up to the river bank, to which
she is just kept up by steam—the plank is run ashore, and the few
passengers who are coming in or out are lighted on their way by the
flames of pine in an iron basket, swinging above the bow by a long
pole. Then we see them vanishing into black darkness up the steps,
or coming down clearer and clearer till they stand in the full blaze of
the beacon which casts dark shadows on the yellow water. The air is
glistening with fire-flies, which dot the darkness with specks and
points of flame, just as sparks fly through the embers of tinder or
half-burnt paper.
Some of the landings were by far more important than others.
There were some, for example, where an iron railroad was worked
down the bank by windlasses for hoisting up goods; others where
the negroes half-naked leaped ashore, and rushing at piles of
firewood, tossed them on board to feed the engine, which, all
uncovered and open to the lower deck, lighted up the darkness by
the glare from the stoke-holes, which cried for ever, “Give, give!” as
the negroes ceaselessly thrust the pine-beams into their hungry
maws. I could understand how easily a steamer can “burn up,” and
how hopeless escape would be under such circumstances. The
whole framework of the vessel is of the lightest resinous pine, so
raw that the turpentine oozes out through the paint; the hull is a
mere shell. If the vessel once caught fire, all that could be done
would be to turn her round, and run her to the bank, in the hope of
holding there long enough to enable the people to escape into the
trees; but if she were not near a landing, many must be lost; as the
bank is steep down, the vessel cannot be run aground; and in some
places the trees are in 8 and 10 feet of water. A few minutes would
suffice to set the vessel in a blaze from stem to stern; and if there
were cotton on board, the bales would burn almost like powder. The
scene at each landing was repeated, with few variations, ten times
till we reached Selma, 110 miles distance, at 11·30 at night.
Selma, which is connected with the Tennessee and Mississippi
rivers by railroad, is built upon a steep, lofty bluff, and the lights in
the windows, and the lofty hotels above us, put me in mind of the
old town of Edinburgh, seen from Princes Street. Beside us there
was a huge storied wharf, so that our passengers could step on
shore from any deck they pleased. Here Mr. Deasy, being attacked
by illness, became alarmed at the idea of continuing his journey
without any opportunity of medical assistance, and went on shore.
May 10th.—The cabin of one of these steamers, in the month of
May, is not favourable to sleep. The wooden beams of the engines
creak and scream “consumedly,” and the great engines themselves
throb as if they would break through their thin, pulse covers of pine,
—and the whistle sounds, and the calliope shrieks out “Dixie”
incessantly. So, when I was up and dressed, breakfast was over, and
I had an opportunity of seeing the slaves on board, male and
female, acting as stewards and stewardesses, at their morning meal,
which they took with much good spirits and decorum. They were
nicely dressed—clean and neat. I was forced to admit to myself that
their Ashantee grandsires and grandmothers, or their Kroo and
Dahomey progenitors were certainly less comfortable and well clad,
and that these slaves had other social advantages, though I could
not recognise the force of the Bishop of Georgia’s assertion, that
from slavery must come the sole hope of, and machinery for, the
evangelisation of Africa. I confess I would not give much for the
influence of the stewards and stewardesses in Christianising the
blacks.
The river, the scenery, and the scenes were just the same as
yesterday’s—high banks, cotton-slides, wooding stations, cane
brakes—and a very miserable negro population, if the specimens of
women and children at the landings fairly represented the mass of
the slaves. They were in strong contrast to the comfortable, well-
dressed domestic slaves on board, and it can well be imagined there
is a wide difference between the classes, and that those condemned
to work in the open fields must suffer exceedingly.
A passenger told us the captain’s story. A number of planters, the
narrator among them, subscribed a thousand dollars each to get up
a vessel for the purpose of running a cargo of slaves, with the
understanding they were to pay so much for the vessel, and so
much per head if she succeeded, and so much if she was taken or
lost. The vessel made her voyage to the coast, was laden with native
Africans, and in due time made her appearance off Mobile. The
collector heard of her, but, oddly enough, the sheriff was not about
at the time, the United States’ Marshal was away, and as the vessel
could not be seen next morning, it was fair to suppose she had gone
up the river, or somewhere or another. But it so happened that
Captain Maher, then commanding a river steamer called the Czar (a
name once very appropriate for the work, but since the serf
emancipation rather out of place), found himself in the
neighbourhood of the brig about nightfall; next morning, indeed, the
Czar was at her moorings in the river; but Captain Maher, began to
grow rich, he had fine negroes fresh run on his land, and bought
fresh acres, and finally built the “Southern Republic.” The planters
asked him for their share of the slaves. Captain Maher laughed
pleasantly; he did not understand what they meant. If he had done
anything wrong, they had their legal remedy. They were completely
beaten; for they could not have recourse to the tribunals in a case
which rendered them liable to capital punishment. And so Captain
Maher, as an act of grace, gave them a few old niggers, and kept the
rest of the cargo.
It was worth while to see the leer with which he listened to this
story about himself, “Wall now! You think them niggers I’ve abord
came from Africa! I’ll show you. Jist come up here, Bully!” A boy of
some twelve years of age, stout, fat, nearly naked, came up to us;
his colour was jet black, his wool close as felt, his cheeks were
marked with regular parallel scars, and his teeth very white, looked
as if they had been filed to a point, his belly was slightly
protuberant, and his chest was marked with tracings of tattoo
marks.
“What’s your name, sir?”
“My name Bully.”
“Where were you born?”
“Me born Sout Karliner, sar!”
“There, you see he wasn’t taken from Africa,” exclaimed the
Captain, knowingly. “I’ve a lot of these black South Caroliny niggers
abord, haven’t I, Bully?”
“Yas, sar.”
“Are you happy, Bully?”
“Yas, sar.”
“Show how you’re happy.”
Here the boy rubbed his stomach, and grinning with delight, said,
“Yummy! yummy! plenty belly full.”
“That’s what I call a real happy feelosophical chap,” quoth the
Captain. “I guess you’ve got a lot in your country can’t pat their
stomachs and say, ‘yummy, yummy, plenty belly full?’”
“Where did he get those marks on his face?”
“Oh, them? Wall, it’s a way them nigger women has of marking
their children to know them; isn’t it, Bully?”
“Yas, sar! me ’spose so!”
“And on his chest!”
“Wall, r’ally I do b’l’eve them’s marks agin the smallpox.”
“Why are his teeth filed?”
“Ah, there now! You’d never have guessed it; Bully done that
himself, for the greater ease of biting his vittels.”
In fact, the lad, and a good many of the hands, were the results
of Captain Maher’s little sail in the Czar.
“We’re obleeged to let ’em in some times to keep up the balance
agin the niggers you run into Canaydy.”
From 1848 to 1852 there were no slaves run; but since the
migrations to Canada and the personal liberty laws, it has been
found profitable to run them. There is a bucolic ferocity about these
Southern people which will stand them good stead in the shock of
battle. How the Spartans would have fought against any barbarians
who came to emancipate their slaves, or the Romans have smitten
those who would manumit slave and creditor together!
To-night, on the lower deck, amid wood faggots, and barrels, a
dance of negroes was arranged by an enthusiast, who desired to
show how “happy they were.” That is the favourite theme of the
Southerners; the gallant Captain Maher becomes quite eloquent
when he points to Bully’s prominent “yummy,” and descants on the
misery of his condition if he had been left to the precarious chances
of obtaining such developments in his native land; then turns a quid,
and, as if uttering some sacred refrain to the universal hymn of the
South, says, “Yes, sir, they’re the happiest people on the face of the
airth!”
There was a fiddler, and also a banjo-player, who played uncouth
music to the clumsiest of dances, which it would be insulting to
compare to the worst Irish jig, and the men with immense gravity
and great effusion of sudor, shuffled, and cut, and heeled and
buckled to each other with an overwhelming solemnity, till the rum-
bottle warmed them up to the lighter graces of the dance, when
they became quite overpowering. “Yes, sir, jist look at them how
they’re enjoying it; they’re the happiest people on the face of the
airth.” When “wooding” and firing up they don’t seem to be in the
possession of the same exquisite felicity.
May 11th.—At early dawn the steamer went its way through a
broad bay of snags bordered with drift-wood, and with steam-
trumpet and calliope announced its arrival at the quay of Mobile,
which presented a fringe of tall warehouses, and shops alongside,
over which were names indicating Scotch, Irish, English, many
Spanish, German, Italian, and French owners, Captain Maher at once
set off to his plantation, and we descended the stories of the walled
castle to the beach, and walked on towards the “Battle House,” so
called from the name of its proprietor, for Mobile has not yet had its
fight like New Orleans. The quays which usually, as we were told,
are lined with stately hulls and a forest of masts, were deserted;
although the port was not actually blockaded, there were squadrons
of the United States ships at Pensacola on the east, and at New
Orleans on the west.
The hotel, a fine building of the American stamp, was the seat of
a Vigilance Committee, and as we put down our names in the book
they were minutely inspected by some gentlemen who came out of
the parlour. It was fortunate they did not find traces of Lincolnism
about us, as it appeared by the papers that they were busy
deporting “Abolitionists” after certain preliminary processes
supposed to—
“Give them a rise, and open their eyes
To a sense of their situation.”
The citizens were busy in drilling, marching, and drum-beating, and
the Confederate flag flew from every spire and steeple. The day was
so hot that it was little more inviting to go out in the sun than it
would be in the dog-days at Malaga, to which, by-the-bye, Mobile
bears some “kinder sorter” resemblance, but, nevertheless, I sallied
forth, and had a drive on a shell road by the head of the bay, where
there were pretty villarettes in charming groves of magnolia, orange-
trees, and lime oaks. Wide streets of similar houses spring out to
meet the country through sandy roads; some worthy of Streatham
or Balham, and all surrounded in such vegetation as Kew might
envy.
Many Mobilians called, and among them the mayor, Mr. Forsyth, in
whom I recognised the most remarkable of the Southern
Commissioners I had met at Washington. Mr. Magee, the acting
British Consul was also good enough to wait upon me, with offers of
any assistance in his power. I hear he has most difficult questions to
deal with, arising out of the claims of distressed British subjects, and
disputed nationality. In the evening the Consul and Dr. Nott, a savant
and physician of Mobile, well known to ethnologists for his work on
the “Types of Mankind,” written conjointly with the late Mr. Gliddon,
dined with me, and I learned from them that, notwithstanding the
intimate commercial relations between Mobile and the great
Northern cities, the people here are of the most ultra-secessionist
doctrines. The wealth and manhood of the city will be devoted to
repel the “Lincolnite mercenaries” to the last.
After dinner we walked through the city, which abounds in oyster
saloons, drinking-houses, lager-bier and wine-shops, and gambling
and dancing places. The market was well worthy of a visit—
something like St. John’s at Liverpool on a Saturday night, crowded
with negroes, mulattoes, quadroons, and mestizos of all sorts,
Spanish, Italian, and French, speaking their own tongues, or a
quaint lingua franca, and dressed in very striking and pretty
costumes. The fruit and vegetable stalls displayed very fine produce,
and some staples, remarkable for novelty, ugliness, and goodness.
After our stroll we went into one of the great oyster saloons, and in
a room up-stairs had opportunity of tasting those great bivalvians in
the form of natural fish puddings, fried in batter, roasted, stewed,
devilled, broiled, and in many other ways, plus raw. I am bound to
observe that the Mobile people ate them as if there was no
blockade, and as though oysters were a specific for political
indigestions and civil wars; a fierce Marseillais are they—living in the
most foreign-looking city I have yet seen in the States. My private
room in the hotel was large, well-lighted with gas, and exceedingly
well furnished in the German fashion, with French pendule and
mirrors. The charge for a private room varies from 1l. to 1l. 5s. a-
day; the bed-room and board are charged separately, from 10s. 6d.
to 12s. 6d. a-day, but meals served in the private room are all
charged extra, and heavily too. Exclusiveness is an aristocratic taste
which must be paid for.
CHAPTER XXVI.
Visit to Forts Gaines and Morgan—War to the knife the cry of the South—The
“State” and the “States”—Bay of Mobile—The forts and their inmates—
Opinions as to an attack on Washington—Rumours of actual war.
Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.
ebooknice.com